Pull kmap_atomic cleanup from Cong Wang.
It's been in -next for a long time, and it gets rid of the (no longer
used) second argument to k[un]map_atomic().
Fix up a few trivial conflicts in various drivers, and do an "evil
merge" to catch some new uses that have come in since Cong's tree.
* 'kmap_atomic' of git://github.com/congwang/linux: (59 commits)
feature-removal-schedule.txt: schedule the deprecated form of kmap_atomic() for removal
highmem: kill all __kmap_atomic() [swarren@nvidia.com: highmem: Fix ARM build break due to __kmap_atomic rename]
drbd: remove the second argument of k[un]map_atomic()
zcache: remove the second argument of k[un]map_atomic()
gma500: remove the second argument of k[un]map_atomic()
dm: remove the second argument of k[un]map_atomic()
tomoyo: remove the second argument of k[un]map_atomic()
sunrpc: remove the second argument of k[un]map_atomic()
rds: remove the second argument of k[un]map_atomic()
net: remove the second argument of k[un]map_atomic()
mm: remove the second argument of k[un]map_atomic()
lib: remove the second argument of k[un]map_atomic()
power: remove the second argument of k[un]map_atomic()
kdb: remove the second argument of k[un]map_atomic()
udf: remove the second argument of k[un]map_atomic()
ubifs: remove the second argument of k[un]map_atomic()
squashfs: remove the second argument of k[un]map_atomic()
reiserfs: remove the second argument of k[un]map_atomic()
ocfs2: remove the second argument of k[un]map_atomic()
ntfs: remove the second argument of k[un]map_atomic()
...
----------------------------
+What: Low Performance USB Block driver ("CONFIG_BLK_DEV_UB")
+When: 3.6
+Why: This driver provides support for USB storage devices like "USB
+ sticks". As of now, it is deactivated in Debian, Fedora and
+ Ubuntu. All current users can switch over to usb-storage
+ (CONFIG_USB_STORAGE) which only drawback is the additional SCSI
+ stack.
+Who: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
++
++----------------------------
++
+ What: kmap_atomic(page, km_type)
+ When: 3.5
+ Why: The old kmap_atomic() with two arguments is deprecated, we only
+ keep it for backward compatibility for few cycles and then drop it.
+ Who: Cong Wang <amwang@redhat.com>
#include <crypto/aes.h>
#include <crypto/cryptd.h>
#include <crypto/ctr.h>
+#include <asm/cpu_device_id.h>
#include <asm/i387.h>
#include <asm/aes.h>
#include <crypto/scatterwalk.h>
one_entry_in_sg = 1;
scatterwalk_start(&src_sg_walk, req->src);
scatterwalk_start(&assoc_sg_walk, req->assoc);
- src = scatterwalk_map(&src_sg_walk, 0);
- assoc = scatterwalk_map(&assoc_sg_walk, 0);
+ src = scatterwalk_map(&src_sg_walk);
+ assoc = scatterwalk_map(&assoc_sg_walk);
dst = src;
if (unlikely(req->src != req->dst)) {
scatterwalk_start(&dst_sg_walk, req->dst);
- dst = scatterwalk_map(&dst_sg_walk, 0);
+ dst = scatterwalk_map(&dst_sg_walk);
}
} else {
* back to the packet. */
if (one_entry_in_sg) {
if (unlikely(req->src != req->dst)) {
- scatterwalk_unmap(dst, 0);
+ scatterwalk_unmap(dst);
scatterwalk_done(&dst_sg_walk, 0, 0);
}
- scatterwalk_unmap(src, 0);
- scatterwalk_unmap(assoc, 0);
+ scatterwalk_unmap(src);
+ scatterwalk_unmap(assoc);
scatterwalk_done(&src_sg_walk, 0, 0);
scatterwalk_done(&assoc_sg_walk, 0, 0);
} else {
one_entry_in_sg = 1;
scatterwalk_start(&src_sg_walk, req->src);
scatterwalk_start(&assoc_sg_walk, req->assoc);
- src = scatterwalk_map(&src_sg_walk, 0);
- assoc = scatterwalk_map(&assoc_sg_walk, 0);
+ src = scatterwalk_map(&src_sg_walk);
+ assoc = scatterwalk_map(&assoc_sg_walk);
dst = src;
if (unlikely(req->src != req->dst)) {
scatterwalk_start(&dst_sg_walk, req->dst);
- dst = scatterwalk_map(&dst_sg_walk, 0);
+ dst = scatterwalk_map(&dst_sg_walk);
}
} else {
if (one_entry_in_sg) {
if (unlikely(req->src != req->dst)) {
- scatterwalk_unmap(dst, 0);
+ scatterwalk_unmap(dst);
scatterwalk_done(&dst_sg_walk, 0, 0);
}
- scatterwalk_unmap(src, 0);
- scatterwalk_unmap(assoc, 0);
+ scatterwalk_unmap(src);
+ scatterwalk_unmap(assoc);
scatterwalk_done(&src_sg_walk, 0, 0);
scatterwalk_done(&assoc_sg_walk, 0, 0);
} else {
};
#endif
+
+static const struct x86_cpu_id aesni_cpu_id[] = {
+ X86_FEATURE_MATCH(X86_FEATURE_AES),
+ {}
+};
+MODULE_DEVICE_TABLE(x86cpu, aesni_cpu_id);
+
static int __init aesni_init(void)
{
int err;
- if (!cpu_has_aes) {
- printk(KERN_INFO "Intel AES-NI instructions are not detected.\n");
+ if (!x86_match_cpu(aesni_cpu_id))
return -ENODEV;
- }
if ((err = crypto_fpu_init()))
goto fpu_err;
eeprom.offset = 0;
data = kmalloc(eeprom.len, GFP_KERNEL);
- if (!data) {
- pr_err("Unable to allocate memory to dump EEPROM data\n");
+ if (!data)
return;
- }
ops->get_eeprom(netdev, &eeprom, data);
(hw->mac_type != e1000_82547))
netdev->hw_features |= NETIF_F_TSO;
+ netdev->priv_flags |= IFF_SUPP_NOFCS;
+
netdev->features |= netdev->hw_features;
netdev->hw_features |= NETIF_F_RXCSUM;
+ netdev->hw_features |= NETIF_F_RXFCS;
if (pci_using_dac) {
netdev->features |= NETIF_F_HIGHDMA;
#define E1000_TX_FLAGS_VLAN 0x00000002
#define E1000_TX_FLAGS_TSO 0x00000004
#define E1000_TX_FLAGS_IPV4 0x00000008
+#define E1000_TX_FLAGS_NO_FCS 0x00000010
#define E1000_TX_FLAGS_VLAN_MASK 0xffff0000
#define E1000_TX_FLAGS_VLAN_SHIFT 16
txd_upper |= (tx_flags & E1000_TX_FLAGS_VLAN_MASK);
}
+ if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
+ txd_lower &= ~(E1000_TXD_CMD_IFCS);
+
i = tx_ring->next_to_use;
while (count--) {
tx_desc->lower.data |= cpu_to_le32(adapter->txd_cmd);
+ /* txd_cmd re-enables FCS, so we'll re-disable it here as desired. */
+ if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
+ tx_desc->lower.data &= ~(cpu_to_le32(E1000_TXD_CMD_IFCS));
+
/* Force memory writes to complete before letting h/w
* know there are new descriptors to fetch. (Only
* applicable for weak-ordered memory model archs,
if (likely(skb->protocol == htons(ETH_P_IP)))
tx_flags |= E1000_TX_FLAGS_IPV4;
+ if (unlikely(skb->no_fcs))
+ tx_flags |= E1000_TX_FLAGS_NO_FCS;
+
count = e1000_tx_map(adapter, tx_ring, skb, first, max_per_txd,
nr_frags, mss);
return NETDEV_TX_OK;
}
+#define NUM_REGS 38 /* 1 based count */
+static void e1000_regdump(struct e1000_adapter *adapter)
+{
+ struct e1000_hw *hw = &adapter->hw;
+ u32 regs[NUM_REGS];
+ u32 *regs_buff = regs;
+ int i = 0;
+
+ static const char * const reg_name[] = {
+ "CTRL", "STATUS",
+ "RCTL", "RDLEN", "RDH", "RDT", "RDTR",
+ "TCTL", "TDBAL", "TDBAH", "TDLEN", "TDH", "TDT",
+ "TIDV", "TXDCTL", "TADV", "TARC0",
+ "TDBAL1", "TDBAH1", "TDLEN1", "TDH1", "TDT1",
+ "TXDCTL1", "TARC1",
+ "CTRL_EXT", "ERT", "RDBAL", "RDBAH",
+ "TDFH", "TDFT", "TDFHS", "TDFTS", "TDFPC",
+ "RDFH", "RDFT", "RDFHS", "RDFTS", "RDFPC"
+ };
+
+ regs_buff[0] = er32(CTRL);
+ regs_buff[1] = er32(STATUS);
+
+ regs_buff[2] = er32(RCTL);
+ regs_buff[3] = er32(RDLEN);
+ regs_buff[4] = er32(RDH);
+ regs_buff[5] = er32(RDT);
+ regs_buff[6] = er32(RDTR);
+
+ regs_buff[7] = er32(TCTL);
+ regs_buff[8] = er32(TDBAL);
+ regs_buff[9] = er32(TDBAH);
+ regs_buff[10] = er32(TDLEN);
+ regs_buff[11] = er32(TDH);
+ regs_buff[12] = er32(TDT);
+ regs_buff[13] = er32(TIDV);
+ regs_buff[14] = er32(TXDCTL);
+ regs_buff[15] = er32(TADV);
+ regs_buff[16] = er32(TARC0);
+
+ regs_buff[17] = er32(TDBAL1);
+ regs_buff[18] = er32(TDBAH1);
+ regs_buff[19] = er32(TDLEN1);
+ regs_buff[20] = er32(TDH1);
+ regs_buff[21] = er32(TDT1);
+ regs_buff[22] = er32(TXDCTL1);
+ regs_buff[23] = er32(TARC1);
+ regs_buff[24] = er32(CTRL_EXT);
+ regs_buff[25] = er32(ERT);
+ regs_buff[26] = er32(RDBAL0);
+ regs_buff[27] = er32(RDBAH0);
+ regs_buff[28] = er32(TDFH);
+ regs_buff[29] = er32(TDFT);
+ regs_buff[30] = er32(TDFHS);
+ regs_buff[31] = er32(TDFTS);
+ regs_buff[32] = er32(TDFPC);
+ regs_buff[33] = er32(RDFH);
+ regs_buff[34] = er32(RDFT);
+ regs_buff[35] = er32(RDFHS);
+ regs_buff[36] = er32(RDFTS);
+ regs_buff[37] = er32(RDFPC);
+
+ pr_info("Register dump\n");
+ for (i = 0; i < NUM_REGS; i++)
+ pr_info("%-15s %08x\n", reg_name[i], regs_buff[i]);
+}
+
+/*
+ * e1000_dump: Print registers, tx ring and rx ring
+ */
+static void e1000_dump(struct e1000_adapter *adapter)
+{
+ /* this code doesn't handle multiple rings */
+ struct e1000_tx_ring *tx_ring = adapter->tx_ring;
+ struct e1000_rx_ring *rx_ring = adapter->rx_ring;
+ int i;
+
+ if (!netif_msg_hw(adapter))
+ return;
+
+ /* Print Registers */
+ e1000_regdump(adapter);
+
+ /*
+ * transmit dump
+ */
+ pr_info("TX Desc ring0 dump\n");
+
+ /* Transmit Descriptor Formats - DEXT[29] is 0 (Legacy) or 1 (Extended)
+ *
+ * Legacy Transmit Descriptor
+ * +--------------------------------------------------------------+
+ * 0 | Buffer Address [63:0] (Reserved on Write Back) |
+ * +--------------------------------------------------------------+
+ * 8 | Special | CSS | Status | CMD | CSO | Length |
+ * +--------------------------------------------------------------+
+ * 63 48 47 36 35 32 31 24 23 16 15 0
+ *
+ * Extended Context Descriptor (DTYP=0x0) for TSO or checksum offload
+ * 63 48 47 40 39 32 31 16 15 8 7 0
+ * +----------------------------------------------------------------+
+ * 0 | TUCSE | TUCS0 | TUCSS | IPCSE | IPCS0 | IPCSS |
+ * +----------------------------------------------------------------+
+ * 8 | MSS | HDRLEN | RSV | STA | TUCMD | DTYP | PAYLEN |
+ * +----------------------------------------------------------------+
+ * 63 48 47 40 39 36 35 32 31 24 23 20 19 0
+ *
+ * Extended Data Descriptor (DTYP=0x1)
+ * +----------------------------------------------------------------+
+ * 0 | Buffer Address [63:0] |
+ * +----------------------------------------------------------------+
+ * 8 | VLAN tag | POPTS | Rsvd | Status | Command | DTYP | DTALEN |
+ * +----------------------------------------------------------------+
+ * 63 48 47 40 39 36 35 32 31 24 23 20 19 0
+ */
+ pr_info("Tc[desc] [Ce CoCsIpceCoS] [MssHlRSCm0Plen] [bi->dma ] leng ntw timestmp bi->skb\n");
+ pr_info("Td[desc] [address 63:0 ] [VlaPoRSCm1Dlen] [bi->dma ] leng ntw timestmp bi->skb\n");
+
+ if (!netif_msg_tx_done(adapter))
+ goto rx_ring_summary;
+
+ for (i = 0; tx_ring->desc && (i < tx_ring->count); i++) {
+ struct e1000_tx_desc *tx_desc = E1000_TX_DESC(*tx_ring, i);
+ struct e1000_buffer *buffer_info = &tx_ring->buffer_info[i];
+ struct my_u { u64 a; u64 b; };
+ struct my_u *u = (struct my_u *)tx_desc;
+ const char *type;
+
+ if (i == tx_ring->next_to_use && i == tx_ring->next_to_clean)
+ type = "NTC/U";
+ else if (i == tx_ring->next_to_use)
+ type = "NTU";
+ else if (i == tx_ring->next_to_clean)
+ type = "NTC";
+ else
+ type = "";
+
+ pr_info("T%c[0x%03X] %016llX %016llX %016llX %04X %3X %016llX %p %s\n",
+ ((le64_to_cpu(u->b) & (1<<20)) ? 'd' : 'c'), i,
+ le64_to_cpu(u->a), le64_to_cpu(u->b),
+ (u64)buffer_info->dma, buffer_info->length,
+ buffer_info->next_to_watch,
+ (u64)buffer_info->time_stamp, buffer_info->skb, type);
+ }
+
+rx_ring_summary:
+ /*
+ * receive dump
+ */
+ pr_info("\nRX Desc ring dump\n");
+
+ /* Legacy Receive Descriptor Format
+ *
+ * +-----------------------------------------------------+
+ * | Buffer Address [63:0] |
+ * +-----------------------------------------------------+
+ * | VLAN Tag | Errors | Status 0 | Packet csum | Length |
+ * +-----------------------------------------------------+
+ * 63 48 47 40 39 32 31 16 15 0
+ */
+ pr_info("R[desc] [address 63:0 ] [vl er S cks ln] [bi->dma ] [bi->skb]\n");
+
+ if (!netif_msg_rx_status(adapter))
+ goto exit;
+
+ for (i = 0; rx_ring->desc && (i < rx_ring->count); i++) {
+ struct e1000_rx_desc *rx_desc = E1000_RX_DESC(*rx_ring, i);
+ struct e1000_buffer *buffer_info = &rx_ring->buffer_info[i];
+ struct my_u { u64 a; u64 b; };
+ struct my_u *u = (struct my_u *)rx_desc;
+ const char *type;
+
+ if (i == rx_ring->next_to_use)
+ type = "NTU";
+ else if (i == rx_ring->next_to_clean)
+ type = "NTC";
+ else
+ type = "";
+
+ pr_info("R[0x%03X] %016llX %016llX %016llX %p %s\n",
+ i, le64_to_cpu(u->a), le64_to_cpu(u->b),
+ (u64)buffer_info->dma, buffer_info->skb, type);
+ } /* for */
+
+ /* dump the descriptor caches */
+ /* rx */
+ pr_info("Rx descriptor cache in 64bit format\n");
+ for (i = 0x6000; i <= 0x63FF ; i += 0x10) {
+ pr_info("R%04X: %08X|%08X %08X|%08X\n",
+ i,
+ readl(adapter->hw.hw_addr + i+4),
+ readl(adapter->hw.hw_addr + i),
+ readl(adapter->hw.hw_addr + i+12),
+ readl(adapter->hw.hw_addr + i+8));
+ }
+ /* tx */
+ pr_info("Tx descriptor cache in 64bit format\n");
+ for (i = 0x7000; i <= 0x73FF ; i += 0x10) {
+ pr_info("T%04X: %08X|%08X %08X|%08X\n",
+ i,
+ readl(adapter->hw.hw_addr + i+4),
+ readl(adapter->hw.hw_addr + i),
+ readl(adapter->hw.hw_addr + i+12),
+ readl(adapter->hw.hw_addr + i+8));
+ }
+exit:
+ return;
+}
+
/**
* e1000_tx_timeout - Respond to a Tx Hang
* @netdev: network interface device structure
if (test_bit(__E1000_DOWN, &adapter->flags))
return;
+ e_err(drv, "Reset adapter\n");
e1000_reinit_safe(adapter);
}
eop,
jiffies,
eop_desc->upper.fields.status);
+ e1000_dump(adapter);
netif_stop_queue(netdev);
}
}
if (length <= copybreak &&
skb_tailroom(skb) >= length) {
u8 *vaddr;
- vaddr = kmap_atomic(buffer_info->page,
- KM_SKB_DATA_SOFTIRQ);
+ vaddr = kmap_atomic(buffer_info->page);
memcpy(skb_tail_pointer(skb), vaddr, length);
- kunmap_atomic(vaddr,
- KM_SKB_DATA_SOFTIRQ);
+ kunmap_atomic(vaddr);
/* re-use the page, so don't erase
* buffer_info->page */
skb_put(skb, length);
((u32)(rx_desc->errors) << 24),
le16_to_cpu(rx_desc->csum), skb);
- pskb_trim(skb, skb->len - 4);
-
- /* probably a little skewed due to removing CRC */
- total_rx_bytes += skb->len;
+ total_rx_bytes += (skb->len - 4); /* don't count FCS */
+ if (likely(!(netdev->features & NETIF_F_RXFCS)))
+ pskb_trim(skb, skb->len - 4);
total_rx_packets++;
/* eth type trans needs skb->data to point to something */
}
}
- /* adjust length to remove Ethernet CRC, this must be
- * done after the TBI_ACCEPT workaround above */
- length -= 4;
-
- /* probably a little skewed due to removing CRC */
- total_rx_bytes += length;
+ total_rx_bytes += (length - 4); /* don't count FCS */
total_rx_packets++;
+ if (likely(!(netdev->features & NETIF_F_RXFCS)))
+ /* adjust length to remove Ethernet CRC, this must be
+ * done after the TBI_ACCEPT workaround above
+ */
+ length -= 4;
+
e1000_check_copybreak(netdev, buffer_info, length, &skb);
skb_put(skb, length);
/*******************************************************************************
Intel PRO/1000 Linux driver
- Copyright(c) 1999 - 2011 Intel Corporation.
+ Copyright(c) 1999 - 2012 Intel Corporation.
This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
#define DRV_EXTRAVERSION "-k"
-#define DRV_VERSION "1.5.1" DRV_EXTRAVERSION
+#define DRV_VERSION "1.9.5" DRV_EXTRAVERSION
char e1000e_driver_name[] = "e1000e";
const char e1000e_driver_version[] = DRV_VERSION;
{E1000_TDFPC, "TDFPC"},
/* List Terminator */
- {}
+ {0, NULL}
};
/*
struct e1000_ring *tx_ring = adapter->tx_ring;
struct e1000_tx_desc *tx_desc;
struct my_u0 {
- u64 a;
- u64 b;
+ __le64 a;
+ __le64 b;
} *u0;
struct e1000_buffer *buffer_info;
struct e1000_ring *rx_ring = adapter->rx_ring;
union e1000_rx_desc_packet_split *rx_desc_ps;
union e1000_rx_desc_extended *rx_desc;
struct my_u1 {
- u64 a;
- u64 b;
- u64 c;
- u64 d;
+ __le64 a;
+ __le64 b;
+ __le64 c;
+ __le64 d;
} *u1;
u32 staterr;
int i = 0;
/* Print Tx Ring Summary */
if (!netdev || !netif_running(netdev))
- goto exit;
+ return;
dev_info(&adapter->pdev->dev, "Tx Ring Summary\n");
pr_info("Queue [NTU] [NTC] [bi(ntc)->dma ] leng ntw timestamp\n");
/* Print Rx Ring */
if (!netif_msg_rx_status(adapter))
- goto exit;
+ return;
dev_info(&adapter->pdev->dev, "Rx Ring Dump\n");
switch (adapter->rx_ps_pages) {
}
}
}
-
-exit:
- return;
}
/**
/**
* e1000_rx_checksum - Receive Checksum Offload
- * @adapter: board private structure
- * @status_err: receive descriptor status and error fields
- * @csum: receive descriptor csum field
- * @sk_buff: socket buffer with received data
+ * @adapter: board private structure
+ * @status_err: receive descriptor status and error fields
+ * @csum: receive descriptor csum field
+ * @sk_buff: socket buffer with received data
**/
static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
- u32 csum, struct sk_buff *skb)
+ __le16 csum, struct sk_buff *skb)
{
u16 status = (u16)status_err;
u8 errors = (u8)(status_err >> 24);
skb_checksum_none_assert(skb);
+ /* Rx checksum disabled */
+ if (!(adapter->netdev->features & NETIF_F_RXCSUM))
+ return;
+
/* Ignore Checksum bit is set */
if (status & E1000_RXD_STAT_IXSM)
return;
+
/* TCP/UDP checksum error bit is set */
if (errors & E1000_RXD_ERR_TCPE) {
/* let the stack verify checksum errors */
* Hardware complements the payload checksum, so we undo it
* and then put the value in host order for further stack use.
*/
- __sum16 sum = (__force __sum16)htons(csum);
+ __sum16 sum = (__force __sum16)swab16((__force u16)csum);
skb->csum = csum_unfold(~sum);
skb->ip_summed = CHECKSUM_COMPLETE;
}
* which has bit 24 set while ME is accessing Host CSR registers, wait
* if it is set and try again a number of times.
**/
-static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, u8 __iomem * tail,
+static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, void __iomem *tail,
unsigned int i)
{
unsigned int j = 0;
return 0;
}
-static void e1000e_update_rdt_wa(struct e1000_adapter *adapter, unsigned int i)
+static void e1000e_update_rdt_wa(struct e1000_ring *rx_ring, unsigned int i)
{
- u8 __iomem *tail = (adapter->hw.hw_addr + adapter->rx_ring->tail);
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct e1000_hw *hw = &adapter->hw;
- if (e1000e_update_tail_wa(hw, tail, i)) {
+ if (e1000e_update_tail_wa(hw, rx_ring->tail, i)) {
u32 rctl = er32(RCTL);
ew32(RCTL, rctl & ~E1000_RCTL_EN);
e_err("ME firmware caused invalid RDT - resetting\n");
}
}
-static void e1000e_update_tdt_wa(struct e1000_adapter *adapter, unsigned int i)
+static void e1000e_update_tdt_wa(struct e1000_ring *tx_ring, unsigned int i)
{
- u8 __iomem *tail = (adapter->hw.hw_addr + adapter->tx_ring->tail);
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct e1000_hw *hw = &adapter->hw;
- if (e1000e_update_tail_wa(hw, tail, i)) {
+ if (e1000e_update_tail_wa(hw, tx_ring->tail, i)) {
u32 tctl = er32(TCTL);
ew32(TCTL, tctl & ~E1000_TCTL_EN);
e_err("ME firmware caused invalid TDT - resetting\n");
/**
* e1000_alloc_rx_buffers - Replace used receive buffers
- * @adapter: address of board private structure
+ * @rx_ring: Rx descriptor ring
**/
-static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
+static void e1000_alloc_rx_buffers(struct e1000_ring *rx_ring,
int cleaned_count, gfp_t gfp)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
- struct e1000_ring *rx_ring = adapter->rx_ring;
union e1000_rx_desc_extended *rx_desc;
struct e1000_buffer *buffer_info;
struct sk_buff *skb;
*/
wmb();
if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
- e1000e_update_rdt_wa(adapter, i);
+ e1000e_update_rdt_wa(rx_ring, i);
else
- writel(i, adapter->hw.hw_addr + rx_ring->tail);
+ writel(i, rx_ring->tail);
}
i++;
if (i == rx_ring->count)
/**
* e1000_alloc_rx_buffers_ps - Replace used receive buffers; packet split
- * @adapter: address of board private structure
+ * @rx_ring: Rx descriptor ring
**/
-static void e1000_alloc_rx_buffers_ps(struct e1000_adapter *adapter,
+static void e1000_alloc_rx_buffers_ps(struct e1000_ring *rx_ring,
int cleaned_count, gfp_t gfp)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
union e1000_rx_desc_packet_split *rx_desc;
- struct e1000_ring *rx_ring = adapter->rx_ring;
struct e1000_buffer *buffer_info;
struct e1000_ps_page *ps_page;
struct sk_buff *skb;
*/
wmb();
if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
- e1000e_update_rdt_wa(adapter, i << 1);
+ e1000e_update_rdt_wa(rx_ring, i << 1);
else
- writel(i << 1,
- adapter->hw.hw_addr + rx_ring->tail);
+ writel(i << 1, rx_ring->tail);
}
i++;
/**
* e1000_alloc_jumbo_rx_buffers - Replace used jumbo receive buffers
- * @adapter: address of board private structure
+ * @rx_ring: Rx descriptor ring
* @cleaned_count: number of buffers to allocate this pass
**/
-static void e1000_alloc_jumbo_rx_buffers(struct e1000_adapter *adapter,
+static void e1000_alloc_jumbo_rx_buffers(struct e1000_ring *rx_ring,
int cleaned_count, gfp_t gfp)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
union e1000_rx_desc_extended *rx_desc;
- struct e1000_ring *rx_ring = adapter->rx_ring;
struct e1000_buffer *buffer_info;
struct sk_buff *skb;
unsigned int i;
* such as IA-64). */
wmb();
if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
- e1000e_update_rdt_wa(adapter, i);
+ e1000e_update_rdt_wa(rx_ring, i);
else
- writel(i, adapter->hw.hw_addr + rx_ring->tail);
+ writel(i, rx_ring->tail);
}
}
+static inline void e1000_rx_hash(struct net_device *netdev, __le32 rss,
+ struct sk_buff *skb)
+{
+ if (netdev->features & NETIF_F_RXHASH)
+ skb->rxhash = le32_to_cpu(rss);
+}
+
/**
- * e1000_clean_rx_irq - Send received data up the network stack; legacy
- * @adapter: board private structure
+ * e1000_clean_rx_irq - Send received data up the network stack
+ * @rx_ring: Rx descriptor ring
*
* the return value indicates whether actual cleaning was done, there
* is no guarantee that everything was cleaned
**/
-static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
- int *work_done, int work_to_do)
+static bool e1000_clean_rx_irq(struct e1000_ring *rx_ring, int *work_done,
+ int work_to_do)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
struct e1000_hw *hw = &adapter->hw;
- struct e1000_ring *rx_ring = adapter->rx_ring;
union e1000_rx_desc_extended *rx_desc, *next_rxd;
struct e1000_buffer *buffer_info, *next_buffer;
u32 length, staterr;
goto next_desc;
}
- if (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) {
+ if (unlikely((staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) &&
+ !(netdev->features & NETIF_F_RXALL))) {
/* recycle */
buffer_info->skb = skb;
goto next_desc;
}
/* adjust length to remove Ethernet CRC */
- if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
- length -= 4;
+ if (!(adapter->flags2 & FLAG2_CRC_STRIPPING)) {
+ /* If configured to store CRC, don't subtract FCS,
+ * but keep the FCS bytes out of the total_rx_bytes
+ * counter
+ */
+ if (netdev->features & NETIF_F_RXFCS)
+ total_rx_bytes -= 4;
+ else
+ length -= 4;
+ }
total_rx_bytes += length;
total_rx_packets++;
/* Receive Checksum Offload */
e1000_rx_checksum(adapter, staterr,
- le16_to_cpu(rx_desc->wb.lower.hi_dword.
- csum_ip.csum), skb);
+ rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
+
+ e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);
e1000_receive_skb(adapter, netdev, skb, staterr,
rx_desc->wb.upper.vlan);
/* return some buffers to hardware, one at a time is too slow */
if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
- adapter->alloc_rx_buf(adapter, cleaned_count,
+ adapter->alloc_rx_buf(rx_ring, cleaned_count,
GFP_ATOMIC);
cleaned_count = 0;
}
cleaned_count = e1000_desc_unused(rx_ring);
if (cleaned_count)
- adapter->alloc_rx_buf(adapter, cleaned_count, GFP_ATOMIC);
+ adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
adapter->total_rx_bytes += total_rx_bytes;
adapter->total_rx_packets += total_rx_packets;
return cleaned;
}
-static void e1000_put_txbuf(struct e1000_adapter *adapter,
- struct e1000_buffer *buffer_info)
+static void e1000_put_txbuf(struct e1000_ring *tx_ring,
+ struct e1000_buffer *buffer_info)
{
+ struct e1000_adapter *adapter = tx_ring->adapter;
+
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
"PHY 1000BASE-T Status <%x>\n"
"PHY Extended Status <%x>\n"
"PCI Status <%x>\n",
- readl(adapter->hw.hw_addr + tx_ring->head),
- readl(adapter->hw.hw_addr + tx_ring->tail),
+ readl(tx_ring->head),
+ readl(tx_ring->tail),
tx_ring->next_to_use,
tx_ring->next_to_clean,
tx_ring->buffer_info[eop].time_stamp,
/**
* e1000_clean_tx_irq - Reclaim resources after transmit completes
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
*
* the return value indicates whether actual cleaning was done, there
* is no guarantee that everything was cleaned
**/
-static bool e1000_clean_tx_irq(struct e1000_adapter *adapter)
+static bool e1000_clean_tx_irq(struct e1000_ring *tx_ring)
{
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct net_device *netdev = adapter->netdev;
struct e1000_hw *hw = &adapter->hw;
- struct e1000_ring *tx_ring = adapter->tx_ring;
struct e1000_tx_desc *tx_desc, *eop_desc;
struct e1000_buffer *buffer_info;
unsigned int i, eop;
}
}
- e1000_put_txbuf(adapter, buffer_info);
+ e1000_put_txbuf(tx_ring, buffer_info);
tx_desc->upper.data = 0;
i++;
/**
* e1000_clean_rx_irq_ps - Send received data up the network stack; packet split
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
*
* the return value indicates whether actual cleaning was done, there
* is no guarantee that everything was cleaned
**/
-static bool e1000_clean_rx_irq_ps(struct e1000_adapter *adapter,
- int *work_done, int work_to_do)
+static bool e1000_clean_rx_irq_ps(struct e1000_ring *rx_ring, int *work_done,
+ int work_to_do)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct e1000_hw *hw = &adapter->hw;
union e1000_rx_desc_packet_split *rx_desc, *next_rxd;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
- struct e1000_ring *rx_ring = adapter->rx_ring;
struct e1000_buffer *buffer_info, *next_buffer;
struct e1000_ps_page *ps_page;
struct sk_buff *skb;
goto next_desc;
}
- if (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) {
+ if (unlikely((staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) &&
+ !(netdev->features & NETIF_F_RXALL))) {
dev_kfree_skb_irq(skb);
goto next_desc;
}
skb_put(skb, length);
{
- /*
- * this looks ugly, but it seems compiler issues make it
- * more efficient than reusing j
- */
- int l1 = le16_to_cpu(rx_desc->wb.upper.length[0]);
-
- /*
- * page alloc/put takes too long and effects small packet
- * throughput, so unsplit small packets and save the alloc/put
- * only valid in softirq (napi) context to call kmap_*
- */
- if (l1 && (l1 <= copybreak) &&
- ((length + l1) <= adapter->rx_ps_bsize0)) {
- u8 *vaddr;
-
- ps_page = &buffer_info->ps_pages[0];
+ /*
+ * this looks ugly, but it seems compiler issues make
+ * it more efficient than reusing j
+ */
+ int l1 = le16_to_cpu(rx_desc->wb.upper.length[0]);
/*
- * there is no documentation about how to call
- * kmap_atomic, so we can't hold the mapping
- * very long
+ * page alloc/put takes too long and effects small
+ * packet throughput, so unsplit small packets and
+ * save the alloc/put only valid in softirq (napi)
+ * context to call kmap_*
*/
- dma_sync_single_for_cpu(&pdev->dev, ps_page->dma,
- PAGE_SIZE, DMA_FROM_DEVICE);
- vaddr = kmap_atomic(ps_page->page);
- memcpy(skb_tail_pointer(skb), vaddr, l1);
- kunmap_atomic(vaddr);
- dma_sync_single_for_device(&pdev->dev, ps_page->dma,
- PAGE_SIZE, DMA_FROM_DEVICE);
-
- /* remove the CRC */
- if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
- l1 -= 4;
-
- skb_put(skb, l1);
- goto copydone;
- } /* if */
+ if (l1 && (l1 <= copybreak) &&
+ ((length + l1) <= adapter->rx_ps_bsize0)) {
+ u8 *vaddr;
+
+ ps_page = &buffer_info->ps_pages[0];
+
+ /*
+ * there is no documentation about how to call
+ * kmap_atomic, so we can't hold the mapping
+ * very long
+ */
+ dma_sync_single_for_cpu(&pdev->dev,
+ ps_page->dma,
+ PAGE_SIZE,
+ DMA_FROM_DEVICE);
- vaddr = kmap_atomic(ps_page->page,
- KM_SKB_DATA_SOFTIRQ);
++ vaddr = kmap_atomic(ps_page->page);
+ memcpy(skb_tail_pointer(skb), vaddr, l1);
- kunmap_atomic(vaddr, KM_SKB_DATA_SOFTIRQ);
++ kunmap_atomic(vaddr);
+ dma_sync_single_for_device(&pdev->dev,
+ ps_page->dma,
+ PAGE_SIZE,
+ DMA_FROM_DEVICE);
+
+ /* remove the CRC */
+ if (!(adapter->flags2 & FLAG2_CRC_STRIPPING)) {
+ if (!(netdev->features & NETIF_F_RXFCS))
+ l1 -= 4;
+ }
+
+ skb_put(skb, l1);
+ goto copydone;
+ } /* if */
}
for (j = 0; j < PS_PAGE_BUFFERS; j++) {
/* strip the ethernet crc, problem is we're using pages now so
* this whole operation can get a little cpu intensive
*/
- if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
- pskb_trim(skb, skb->len - 4);
+ if (!(adapter->flags2 & FLAG2_CRC_STRIPPING)) {
+ if (!(netdev->features & NETIF_F_RXFCS))
+ pskb_trim(skb, skb->len - 4);
+ }
copydone:
total_rx_bytes += skb->len;
total_rx_packets++;
- e1000_rx_checksum(adapter, staterr, le16_to_cpu(
- rx_desc->wb.lower.hi_dword.csum_ip.csum), skb);
+ e1000_rx_checksum(adapter, staterr,
+ rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
+
+ e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);
if (rx_desc->wb.upper.header_status &
cpu_to_le16(E1000_RXDPS_HDRSTAT_HDRSP))
/* return some buffers to hardware, one at a time is too slow */
if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
- adapter->alloc_rx_buf(adapter, cleaned_count,
+ adapter->alloc_rx_buf(rx_ring, cleaned_count,
GFP_ATOMIC);
cleaned_count = 0;
}
cleaned_count = e1000_desc_unused(rx_ring);
if (cleaned_count)
- adapter->alloc_rx_buf(adapter, cleaned_count, GFP_ATOMIC);
+ adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
adapter->total_rx_bytes += total_rx_bytes;
adapter->total_rx_packets += total_rx_packets;
* the return value indicates whether actual cleaning was done, there
* is no guarantee that everything was cleaned
**/
-
-static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
- int *work_done, int work_to_do)
+static bool e1000_clean_jumbo_rx_irq(struct e1000_ring *rx_ring, int *work_done,
+ int work_to_do)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
- struct e1000_ring *rx_ring = adapter->rx_ring;
union e1000_rx_desc_extended *rx_desc, *next_rxd;
struct e1000_buffer *buffer_info, *next_buffer;
u32 length, staterr;
/* errors is only valid for DD + EOP descriptors */
if (unlikely((staterr & E1000_RXD_STAT_EOP) &&
- (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK))) {
+ ((staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) &&
+ !(netdev->features & NETIF_F_RXALL)))) {
/* recycle both page and skb */
buffer_info->skb = skb;
/* an error means any chain goes out the window too */
if (length <= copybreak &&
skb_tailroom(skb) >= length) {
u8 *vaddr;
- vaddr = kmap_atomic(buffer_info->page,
- KM_SKB_DATA_SOFTIRQ);
+ vaddr = kmap_atomic(buffer_info->page);
memcpy(skb_tail_pointer(skb), vaddr,
length);
- kunmap_atomic(vaddr,
- KM_SKB_DATA_SOFTIRQ);
+ kunmap_atomic(vaddr);
/* re-use the page, so don't erase
* buffer_info->page */
skb_put(skb, length);
/* Receive Checksum Offload XXX recompute due to CRC strip? */
e1000_rx_checksum(adapter, staterr,
- le16_to_cpu(rx_desc->wb.lower.hi_dword.
- csum_ip.csum), skb);
+ rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
+
+ e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);
/* probably a little skewed due to removing CRC */
total_rx_bytes += skb->len;
/* return some buffers to hardware, one at a time is too slow */
if (unlikely(cleaned_count >= E1000_RX_BUFFER_WRITE)) {
- adapter->alloc_rx_buf(adapter, cleaned_count,
+ adapter->alloc_rx_buf(rx_ring, cleaned_count,
GFP_ATOMIC);
cleaned_count = 0;
}
cleaned_count = e1000_desc_unused(rx_ring);
if (cleaned_count)
- adapter->alloc_rx_buf(adapter, cleaned_count, GFP_ATOMIC);
+ adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
adapter->total_rx_bytes += total_rx_bytes;
adapter->total_rx_packets += total_rx_packets;
/**
* e1000_clean_rx_ring - Free Rx Buffers per Queue
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
**/
-static void e1000_clean_rx_ring(struct e1000_adapter *adapter)
+static void e1000_clean_rx_ring(struct e1000_ring *rx_ring)
{
- struct e1000_ring *rx_ring = adapter->rx_ring;
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct e1000_buffer *buffer_info;
struct e1000_ps_page *ps_page;
struct pci_dev *pdev = adapter->pdev;
rx_ring->next_to_use = 0;
adapter->flags2 &= ~FLAG2_IS_DISCARDING;
- writel(0, adapter->hw.hw_addr + rx_ring->head);
- writel(0, adapter->hw.hw_addr + rx_ring->tail);
+ writel(0, rx_ring->head);
+ writel(0, rx_ring->tail);
}
static void e1000e_downshift_workaround(struct work_struct *work)
*/
if (icr & E1000_ICR_LSC) {
- hw->mac.get_link_status = 1;
+ hw->mac.get_link_status = true;
/*
* ICH8 workaround-- Call gig speed drop workaround on cable
* disconnect (LSC) before accessing any PHY registers
*/
if (icr & E1000_ICR_LSC) {
- hw->mac.get_link_status = 1;
+ hw->mac.get_link_status = true;
/*
* ICH8 workaround-- Call gig speed drop workaround on cable
* disconnect (LSC) before accessing any PHY registers
if (icr & E1000_ICR_OTHER) {
if (!(icr & E1000_ICR_LSC))
goto no_link_interrupt;
- hw->mac.get_link_status = 1;
+ hw->mac.get_link_status = true;
/* guard against interrupt when we're going down */
if (!test_bit(__E1000_DOWN, &adapter->state))
mod_timer(&adapter->watchdog_timer, jiffies + 1);
adapter->total_tx_bytes = 0;
adapter->total_tx_packets = 0;
- if (!e1000_clean_tx_irq(adapter))
+ if (!e1000_clean_tx_irq(tx_ring))
/* Ring was not completely cleaned, so fire another interrupt */
ew32(ICS, tx_ring->ims_val);
{
struct net_device *netdev = data;
struct e1000_adapter *adapter = netdev_priv(netdev);
+ struct e1000_ring *rx_ring = adapter->rx_ring;
/* Write the ITR value calculated at the end of the
* previous interrupt.
*/
- if (adapter->rx_ring->set_itr) {
- writel(1000000000 / (adapter->rx_ring->itr_val * 256),
- adapter->hw.hw_addr + adapter->rx_ring->itr_register);
- adapter->rx_ring->set_itr = 0;
+ if (rx_ring->set_itr) {
+ writel(1000000000 / (rx_ring->itr_val * 256),
+ rx_ring->itr_register);
+ rx_ring->set_itr = 0;
}
if (napi_schedule_prep(&adapter->napi)) {
adapter->eiac_mask |= rx_ring->ims_val;
if (rx_ring->itr_val)
writel(1000000000 / (rx_ring->itr_val * 256),
- hw->hw_addr + rx_ring->itr_register);
+ rx_ring->itr_register);
else
- writel(1, hw->hw_addr + rx_ring->itr_register);
+ writel(1, rx_ring->itr_register);
ivar = E1000_IVAR_INT_ALLOC_VALID | vector;
/* Configure Tx vector */
vector++;
if (tx_ring->itr_val)
writel(1000000000 / (tx_ring->itr_val * 256),
- hw->hw_addr + tx_ring->itr_register);
+ tx_ring->itr_register);
else
- writel(1, hw->hw_addr + tx_ring->itr_register);
+ writel(1, tx_ring->itr_register);
adapter->eiac_mask |= tx_ring->ims_val;
ivar |= ((E1000_IVAR_INT_ALLOC_VALID | vector) << 8);
e1000_intr_msix_rx, 0, adapter->rx_ring->name,
netdev);
if (err)
- goto out;
- adapter->rx_ring->itr_register = E1000_EITR_82574(vector);
+ return err;
+ adapter->rx_ring->itr_register = adapter->hw.hw_addr +
+ E1000_EITR_82574(vector);
adapter->rx_ring->itr_val = adapter->itr;
vector++;
e1000_intr_msix_tx, 0, adapter->tx_ring->name,
netdev);
if (err)
- goto out;
- adapter->tx_ring->itr_register = E1000_EITR_82574(vector);
+ return err;
+ adapter->tx_ring->itr_register = adapter->hw.hw_addr +
+ E1000_EITR_82574(vector);
adapter->tx_ring->itr_val = adapter->itr;
vector++;
err = request_irq(adapter->msix_entries[vector].vector,
e1000_msix_other, 0, netdev->name, netdev);
if (err)
- goto out;
+ return err;
e1000_configure_msix(adapter);
+
return 0;
-out:
- return err;
}
/**
/**
* e1000e_setup_tx_resources - allocate Tx resources (Descriptors)
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
*
* Return 0 on success, negative on failure
**/
-int e1000e_setup_tx_resources(struct e1000_adapter *adapter)
+int e1000e_setup_tx_resources(struct e1000_ring *tx_ring)
{
- struct e1000_ring *tx_ring = adapter->tx_ring;
+ struct e1000_adapter *adapter = tx_ring->adapter;
int err = -ENOMEM, size;
size = sizeof(struct e1000_buffer) * tx_ring->count;
/**
* e1000e_setup_rx_resources - allocate Rx resources (Descriptors)
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
*
* Returns 0 on success, negative on failure
**/
-int e1000e_setup_rx_resources(struct e1000_adapter *adapter)
+int e1000e_setup_rx_resources(struct e1000_ring *rx_ring)
{
- struct e1000_ring *rx_ring = adapter->rx_ring;
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct e1000_buffer *buffer_info;
int i, size, desc_len, err = -ENOMEM;
/**
* e1000_clean_tx_ring - Free Tx Buffers
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
**/
-static void e1000_clean_tx_ring(struct e1000_adapter *adapter)
+static void e1000_clean_tx_ring(struct e1000_ring *tx_ring)
{
- struct e1000_ring *tx_ring = adapter->tx_ring;
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct e1000_buffer *buffer_info;
unsigned long size;
unsigned int i;
for (i = 0; i < tx_ring->count; i++) {
buffer_info = &tx_ring->buffer_info[i];
- e1000_put_txbuf(adapter, buffer_info);
+ e1000_put_txbuf(tx_ring, buffer_info);
}
netdev_reset_queue(adapter->netdev);
tx_ring->next_to_use = 0;
tx_ring->next_to_clean = 0;
- writel(0, adapter->hw.hw_addr + tx_ring->head);
- writel(0, adapter->hw.hw_addr + tx_ring->tail);
+ writel(0, tx_ring->head);
+ writel(0, tx_ring->tail);
}
/**
* e1000e_free_tx_resources - Free Tx Resources per Queue
- * @adapter: board private structure
+ * @tx_ring: Tx descriptor ring
*
* Free all transmit software resources
**/
-void e1000e_free_tx_resources(struct e1000_adapter *adapter)
+void e1000e_free_tx_resources(struct e1000_ring *tx_ring)
{
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct pci_dev *pdev = adapter->pdev;
- struct e1000_ring *tx_ring = adapter->tx_ring;
- e1000_clean_tx_ring(adapter);
+ e1000_clean_tx_ring(tx_ring);
vfree(tx_ring->buffer_info);
tx_ring->buffer_info = NULL;
/**
* e1000e_free_rx_resources - Free Rx Resources
- * @adapter: board private structure
+ * @rx_ring: Rx descriptor ring
*
* Free all receive software resources
**/
-
-void e1000e_free_rx_resources(struct e1000_adapter *adapter)
+void e1000e_free_rx_resources(struct e1000_ring *rx_ring)
{
+ struct e1000_adapter *adapter = rx_ring->adapter;
struct pci_dev *pdev = adapter->pdev;
- struct e1000_ring *rx_ring = adapter->rx_ring;
int i;
- e1000_clean_rx_ring(adapter);
+ e1000_clean_rx_ring(rx_ring);
for (i = 0; i < rx_ring->count; i++)
kfree(rx_ring->buffer_info[i].ps_pages);
unsigned int retval = itr_setting;
if (packets == 0)
- goto update_itr_done;
+ return itr_setting;
switch (itr_setting) {
case lowest_latency:
break;
}
-update_itr_done:
return retval;
}
**/
static int __devinit e1000_alloc_queues(struct e1000_adapter *adapter)
{
- adapter->tx_ring = kzalloc(sizeof(struct e1000_ring), GFP_KERNEL);
+ int size = sizeof(struct e1000_ring);
+
+ adapter->tx_ring = kzalloc(size, GFP_KERNEL);
if (!adapter->tx_ring)
goto err;
+ adapter->tx_ring->count = adapter->tx_ring_count;
+ adapter->tx_ring->adapter = adapter;
- adapter->rx_ring = kzalloc(sizeof(struct e1000_ring), GFP_KERNEL);
+ adapter->rx_ring = kzalloc(size, GFP_KERNEL);
if (!adapter->rx_ring)
goto err;
+ adapter->rx_ring->count = adapter->rx_ring_count;
+ adapter->rx_ring->adapter = adapter;
return 0;
err:
!(adapter->rx_ring->ims_val & adapter->tx_ring->ims_val))
goto clean_rx;
- tx_cleaned = e1000_clean_tx_irq(adapter);
+ tx_cleaned = e1000_clean_tx_irq(adapter->tx_ring);
clean_rx:
- adapter->clean_rx(adapter, &work_done, budget);
+ adapter->clean_rx(adapter->rx_ring, &work_done, budget);
if (!tx_cleaned)
work_done = budget;
struct e1000_hw *hw = &adapter->hw;
struct e1000_ring *tx_ring = adapter->tx_ring;
u64 tdba;
- u32 tdlen, tctl, tipg, tarc;
- u32 ipgr1, ipgr2;
+ u32 tdlen, tarc;
/* Setup the HW Tx Head and Tail descriptor pointers */
tdba = tx_ring->dma;
ew32(TDLEN, tdlen);
ew32(TDH, 0);
ew32(TDT, 0);
- tx_ring->head = E1000_TDH;
- tx_ring->tail = E1000_TDT;
-
- /* Set the default values for the Tx Inter Packet Gap timer */
- tipg = DEFAULT_82543_TIPG_IPGT_COPPER; /* 8 */
- ipgr1 = DEFAULT_82543_TIPG_IPGR1; /* 8 */
- ipgr2 = DEFAULT_82543_TIPG_IPGR2; /* 6 */
-
- if (adapter->flags & FLAG_TIPG_MEDIUM_FOR_80003ESLAN)
- ipgr2 = DEFAULT_80003ES2LAN_TIPG_IPGR2; /* 7 */
-
- tipg |= ipgr1 << E1000_TIPG_IPGR1_SHIFT;
- tipg |= ipgr2 << E1000_TIPG_IPGR2_SHIFT;
- ew32(TIPG, tipg);
+ tx_ring->head = adapter->hw.hw_addr + E1000_TDH;
+ tx_ring->tail = adapter->hw.hw_addr + E1000_TDT;
/* Set the Tx Interrupt Delay register */
ew32(TIDV, adapter->tx_int_delay);
*/
txdctl |= E1000_TXDCTL_DMA_BURST_ENABLE;
ew32(TXDCTL(0), txdctl);
- /* erratum work around: set txdctl the same for both queues */
- ew32(TXDCTL(1), txdctl);
}
-
- /* Program the Transmit Control Register */
- tctl = er32(TCTL);
- tctl &= ~E1000_TCTL_CT;
- tctl |= E1000_TCTL_PSP | E1000_TCTL_RTLC |
- (E1000_COLLISION_THRESHOLD << E1000_CT_SHIFT);
+ /* erratum work around: set txdctl the same for both queues */
+ ew32(TXDCTL(1), er32(TXDCTL(0)));
if (adapter->flags & FLAG_TARC_SPEED_MODE_BIT) {
tarc = er32(TARC(0));
/* enable Report Status bit */
adapter->txd_cmd |= E1000_TXD_CMD_RS;
- ew32(TCTL, tctl);
-
- e1000e_config_collision_dist(hw);
+ hw->mac.ops.config_collision_dist(hw);
}
/**
* per packet.
*/
pages = PAGE_USE_COUNT(adapter->netdev->mtu);
- if (!(adapter->flags & FLAG_HAS_ERT) && (pages <= 3) &&
- (PAGE_SIZE <= 16384) && (rctl & E1000_RCTL_LPE))
+ if ((pages <= 3) && (PAGE_SIZE <= 16384) && (rctl & E1000_RCTL_LPE))
adapter->rx_ps_pages = pages;
else
adapter->rx_ps_pages = 0;
ew32(PSRCTL, psrctl);
}
+ /* This is useful for sniffing bad packets. */
+ if (adapter->netdev->features & NETIF_F_RXALL) {
+ /* UPE and MPE will be handled by normal PROMISC logic
+ * in e1000e_set_rx_mode */
+ rctl |= (E1000_RCTL_SBP | /* Receive bad packets */
+ E1000_RCTL_BAM | /* RX All Bcast Pkts */
+ E1000_RCTL_PMCF); /* RX All MAC Ctrl Pkts */
+
+ rctl &= ~(E1000_RCTL_VFE | /* Disable VLAN filter */
+ E1000_RCTL_DPF | /* Allow filtered pause */
+ E1000_RCTL_CFIEN); /* Dis VLAN CFIEN Filter */
+ /* Do not mess with E1000_CTRL_VME, it affects transmit as well,
+ * and that breaks VLANs.
+ */
+ }
+
ew32(RFCTL, rfctl);
ew32(RCTL, rctl);
/* just started the receive unit, no need to restart */
ew32(RDLEN, rdlen);
ew32(RDH, 0);
ew32(RDT, 0);
- rx_ring->head = E1000_RDH;
- rx_ring->tail = E1000_RDT;
+ rx_ring->head = adapter->hw.hw_addr + E1000_RDH;
+ rx_ring->tail = adapter->hw.hw_addr + E1000_RDT;
/* Enable Receive Checksum Offload for TCP and UDP */
rxcsum = er32(RXCSUM);
}
ew32(RXCSUM, rxcsum);
- /*
- * Enable early receives on supported devices, only takes effect when
- * packet size is equal or larger than the specified value (in 8 byte
- * units), e.g. using jumbo frames when setting to E1000_ERT_2048
- */
- if ((adapter->flags & FLAG_HAS_ERT) ||
- (adapter->hw.mac.type == e1000_pch2lan)) {
+ if (adapter->hw.mac.type == e1000_pch2lan) {
+ /*
+ * With jumbo frames, excessive C-state transition
+ * latencies result in dropped transactions.
+ */
if (adapter->netdev->mtu > ETH_DATA_LEN) {
u32 rxdctl = er32(RXDCTL(0));
ew32(RXDCTL(0), rxdctl | 0x3);
- if (adapter->flags & FLAG_HAS_ERT)
- ew32(ERT, E1000_ERT_2048 | (1 << 13));
- /*
- * With jumbo frames and early-receive enabled,
- * excessive C-state transition latencies result in
- * dropped transactions.
- */
pm_qos_update_request(&adapter->netdev->pm_qos_req, 55);
} else {
pm_qos_update_request(&adapter->netdev->pm_qos_req,
e1000e_vlan_filter_disable(adapter);
} else {
int count;
+
if (netdev->flags & IFF_ALLMULTI) {
rctl |= E1000_RCTL_MPE;
} else {
e1000e_vlan_strip_disable(adapter);
}
+static void e1000e_setup_rss_hash(struct e1000_adapter *adapter)
+{
+ struct e1000_hw *hw = &adapter->hw;
+ u32 mrqc, rxcsum;
+ int i;
+ static const u32 rsskey[10] = {
+ 0xda565a6d, 0xc20e5b25, 0x3d256741, 0xb08fa343, 0xcb2bcad0,
+ 0xb4307bae, 0xa32dcb77, 0x0cf23080, 0x3bb7426a, 0xfa01acbe
+ };
+
+ /* Fill out hash function seed */
+ for (i = 0; i < 10; i++)
+ ew32(RSSRK(i), rsskey[i]);
+
+ /* Direct all traffic to queue 0 */
+ for (i = 0; i < 32; i++)
+ ew32(RETA(i), 0);
+
+ /*
+ * Disable raw packet checksumming so that RSS hash is placed in
+ * descriptor on writeback.
+ */
+ rxcsum = er32(RXCSUM);
+ rxcsum |= E1000_RXCSUM_PCSD;
+
+ ew32(RXCSUM, rxcsum);
+
+ mrqc = (E1000_MRQC_RSS_FIELD_IPV4 |
+ E1000_MRQC_RSS_FIELD_IPV4_TCP |
+ E1000_MRQC_RSS_FIELD_IPV6 |
+ E1000_MRQC_RSS_FIELD_IPV6_TCP |
+ E1000_MRQC_RSS_FIELD_IPV6_TCP_EX);
+
+ ew32(MRQC, mrqc);
+}
+
/**
* e1000_configure - configure the hardware for Rx and Tx
* @adapter: private board structure
**/
static void e1000_configure(struct e1000_adapter *adapter)
{
+ struct e1000_ring *rx_ring = adapter->rx_ring;
+
e1000e_set_rx_mode(adapter->netdev);
e1000_restore_vlan(adapter);
e1000_init_manageability_pt(adapter);
e1000_configure_tx(adapter);
+
+ if (adapter->netdev->features & NETIF_F_RXHASH)
+ e1000e_setup_rss_hash(adapter);
e1000_setup_rctl(adapter);
e1000_configure_rx(adapter);
- adapter->alloc_rx_buf(adapter, e1000_desc_unused(adapter->rx_ring),
- GFP_KERNEL);
+ adapter->alloc_rx_buf(rx_ring, e1000_desc_unused(rx_ring), GFP_KERNEL);
}
/**
* if short on Rx space, Rx wins and must trump Tx
* adjustment or use Early Receive if available
*/
- if ((pba < min_rx_space) &&
- (!(adapter->flags & FLAG_HAS_ERT)))
- /* ERT enabled in e1000_configure_rx */
+ if (pba < min_rx_space)
pba = min_rx_space;
}
* (or the size used for early receive) above it in the Rx FIFO.
* Set it to the lower of:
* - 90% of the Rx FIFO size, and
- * - the full Rx FIFO size minus the early receive size (for parts
- * with ERT support assuming ERT set to E1000_ERT_2048), or
* - the full Rx FIFO size minus one full frame
*/
if (adapter->flags & FLAG_DISABLE_FC_PAUSE_TIME)
fc->pause_time = 0xFFFF;
else
fc->pause_time = E1000_FC_PAUSE_TIME;
- fc->send_xon = 1;
+ fc->send_xon = true;
fc->current_mode = fc->requested_mode;
switch (hw->mac.type) {
+ case e1000_ich9lan:
+ case e1000_ich10lan:
+ if (adapter->netdev->mtu > ETH_DATA_LEN) {
+ pba = 14;
+ ew32(PBA, pba);
+ fc->high_water = 0x2800;
+ fc->low_water = fc->high_water - 8;
+ break;
+ }
+ /* fall-through */
default:
- if ((adapter->flags & FLAG_HAS_ERT) &&
- (adapter->netdev->mtu > ETH_DATA_LEN))
- hwm = min(((pba << 10) * 9 / 10),
- ((pba << 10) - (E1000_ERT_2048 << 3)));
- else
- hwm = min(((pba << 10) * 9 / 10),
- ((pba << 10) - adapter->max_frame_size));
+ hwm = min(((pba << 10) * 9 / 10),
+ ((pba << 10) - adapter->max_frame_size));
fc->high_water = hwm & E1000_FCRTH_RTH; /* 8-byte granularity */
fc->low_water = fc->high_water - 8;
/*
* Disable Adaptive Interrupt Moderation if 2 full packets cannot
- * fit in receive buffer and early-receive not supported.
+ * fit in receive buffer.
*/
if (adapter->itr_setting & 0x3) {
- if (((adapter->max_frame_size * 2) > (pba << 10)) &&
- !(adapter->flags & FLAG_HAS_ERT)) {
+ if ((adapter->max_frame_size * 2) > (pba << 10)) {
if (!(adapter->flags2 & FLAG2_DISABLE_AIM)) {
dev_info(&adapter->pdev->dev,
"Interrupt Throttle Rate turned off\n");
spin_unlock(&adapter->stats64_lock);
e1000e_flush_descriptors(adapter);
- e1000_clean_tx_ring(adapter);
- e1000_clean_rx_ring(adapter);
+ e1000_clean_tx_ring(adapter->tx_ring);
+ e1000_clean_rx_ring(adapter->rx_ring);
adapter->link_speed = 0;
adapter->link_duplex = 0;
adapter->rx_ps_bsize0 = 128;
adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
+ adapter->tx_ring_count = E1000_DEFAULT_TXD;
+ adapter->rx_ring_count = E1000_DEFAULT_RXD;
spin_lock_init(&adapter->stats64_lock);
if (adapter->flags & FLAG_MSI_TEST_FAILED) {
adapter->int_mode = E1000E_INT_MODE_LEGACY;
e_info("MSI interrupt test failed, using legacy interrupt.\n");
- } else
+ } else {
e_dbg("MSI interrupt test succeeded!\n");
+ }
free_irq(adapter->pdev->irq, netdev);
pci_disable_msi(adapter->pdev);
netif_carrier_off(netdev);
/* allocate transmit descriptors */
- err = e1000e_setup_tx_resources(adapter);
+ err = e1000e_setup_tx_resources(adapter->tx_ring);
if (err)
goto err_setup_tx;
/* allocate receive descriptors */
- err = e1000e_setup_rx_resources(adapter);
+ err = e1000e_setup_rx_resources(adapter->rx_ring);
if (err)
goto err_setup_rx;
E1000_MNG_DHCP_COOKIE_STATUS_VLAN))
e1000_update_mng_vlan(adapter);
- /* DMA latency requirement to workaround early-receive/jumbo issue */
- if ((adapter->flags & FLAG_HAS_ERT) ||
- (adapter->hw.mac.type == e1000_pch2lan))
+ /* DMA latency requirement to workaround jumbo issue */
+ if (adapter->hw.mac.type == e1000_pch2lan)
pm_qos_add_request(&adapter->netdev->pm_qos_req,
PM_QOS_CPU_DMA_LATENCY,
PM_QOS_DEFAULT_VALUE);
err_req_irq:
e1000e_release_hw_control(adapter);
e1000_power_down_phy(adapter);
- e1000e_free_rx_resources(adapter);
+ e1000e_free_rx_resources(adapter->rx_ring);
err_setup_rx:
- e1000e_free_tx_resources(adapter);
+ e1000e_free_tx_resources(adapter->tx_ring);
err_setup_tx:
e1000e_reset(adapter);
pm_runtime_put_sync(&pdev->dev);
}
e1000_power_down_phy(adapter);
- e1000e_free_tx_resources(adapter);
- e1000e_free_rx_resources(adapter);
+ e1000e_free_tx_resources(adapter->tx_ring);
+ e1000e_free_rx_resources(adapter->rx_ring);
/*
* kill manageability vlan ID if supported, but not if a vlan with
!test_bit(__E1000_TESTING, &adapter->state))
e1000e_release_hw_control(adapter);
- if ((adapter->flags & FLAG_HAS_ERT) ||
- (adapter->hw.mac.type == e1000_pch2lan))
+ if (adapter->hw.mac.type == e1000_pch2lan)
pm_qos_remove_request(&adapter->netdev->pm_qos_req);
pm_runtime_put_sync(&pdev->dev);
#define E1000_TX_FLAGS_VLAN 0x00000002
#define E1000_TX_FLAGS_TSO 0x00000004
#define E1000_TX_FLAGS_IPV4 0x00000008
+#define E1000_TX_FLAGS_NO_FCS 0x00000010
#define E1000_TX_FLAGS_VLAN_MASK 0xffff0000
#define E1000_TX_FLAGS_VLAN_SHIFT 16
-static int e1000_tso(struct e1000_adapter *adapter,
- struct sk_buff *skb)
+static int e1000_tso(struct e1000_ring *tx_ring, struct sk_buff *skb)
{
- struct e1000_ring *tx_ring = adapter->tx_ring;
struct e1000_context_desc *context_desc;
struct e1000_buffer *buffer_info;
unsigned int i;
return 1;
}
-static bool e1000_tx_csum(struct e1000_adapter *adapter, struct sk_buff *skb)
+static bool e1000_tx_csum(struct e1000_ring *tx_ring, struct sk_buff *skb)
{
- struct e1000_ring *tx_ring = adapter->tx_ring;
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct e1000_context_desc *context_desc;
struct e1000_buffer *buffer_info;
unsigned int i;
#define E1000_MAX_PER_TXD 8192
#define E1000_MAX_TXD_PWR 12
-static int e1000_tx_map(struct e1000_adapter *adapter,
- struct sk_buff *skb, unsigned int first,
- unsigned int max_per_txd, unsigned int nr_frags,
- unsigned int mss)
+static int e1000_tx_map(struct e1000_ring *tx_ring, struct sk_buff *skb,
+ unsigned int first, unsigned int max_per_txd,
+ unsigned int nr_frags, unsigned int mss)
{
- struct e1000_ring *tx_ring = adapter->tx_ring;
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct pci_dev *pdev = adapter->pdev;
struct e1000_buffer *buffer_info;
unsigned int len = skb_headlen(skb);
i += tx_ring->count;
i--;
buffer_info = &tx_ring->buffer_info[i];
- e1000_put_txbuf(adapter, buffer_info);
+ e1000_put_txbuf(tx_ring, buffer_info);
}
return 0;
}
-static void e1000_tx_queue(struct e1000_adapter *adapter,
- int tx_flags, int count)
+static void e1000_tx_queue(struct e1000_ring *tx_ring, int tx_flags, int count)
{
- struct e1000_ring *tx_ring = adapter->tx_ring;
+ struct e1000_adapter *adapter = tx_ring->adapter;
struct e1000_tx_desc *tx_desc = NULL;
struct e1000_buffer *buffer_info;
u32 txd_upper = 0, txd_lower = E1000_TXD_CMD_IFCS;
txd_upper |= (tx_flags & E1000_TX_FLAGS_VLAN_MASK);
}
+ if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
+ txd_lower &= ~(E1000_TXD_CMD_IFCS);
+
i = tx_ring->next_to_use;
do {
tx_desc->lower.data |= cpu_to_le32(adapter->txd_cmd);
+ /* txd_cmd re-enables FCS, so we'll re-disable it here as desired. */
+ if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
+ tx_desc->lower.data &= ~(cpu_to_le32(E1000_TXD_CMD_IFCS));
+
/*
* Force memory writes to complete before letting h/w
* know there are new descriptors to fetch. (Only
tx_ring->next_to_use = i;
if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
- e1000e_update_tdt_wa(adapter, i);
+ e1000e_update_tdt_wa(tx_ring, i);
else
- writel(i, adapter->hw.hw_addr + tx_ring->tail);
+ writel(i, tx_ring->tail);
/*
* we need this if more than one processor can write to our tail
return 0;
}
-static int __e1000_maybe_stop_tx(struct net_device *netdev, int size)
+static int __e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
{
- struct e1000_adapter *adapter = netdev_priv(netdev);
+ struct e1000_adapter *adapter = tx_ring->adapter;
- netif_stop_queue(netdev);
+ netif_stop_queue(adapter->netdev);
/*
* Herbert's original patch had:
* smp_mb__after_netif_stop_queue();
* We need to check again in a case another CPU has just
* made room available.
*/
- if (e1000_desc_unused(adapter->tx_ring) < size)
+ if (e1000_desc_unused(tx_ring) < size)
return -EBUSY;
/* A reprieve! */
- netif_start_queue(netdev);
+ netif_start_queue(adapter->netdev);
++adapter->restart_queue;
return 0;
}
-static int e1000_maybe_stop_tx(struct net_device *netdev, int size)
+static int e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
{
- struct e1000_adapter *adapter = netdev_priv(netdev);
-
- if (e1000_desc_unused(adapter->tx_ring) >= size)
+ if (e1000_desc_unused(tx_ring) >= size)
return 0;
- return __e1000_maybe_stop_tx(netdev, size);
+ return __e1000_maybe_stop_tx(tx_ring, size);
}
-#define TXD_USE_COUNT(S, X) (((S) >> (X)) + 1 )
+#define TXD_USE_COUNT(S, X) (((S) >> (X)) + 1)
static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
struct net_device *netdev)
{
if (skb->data_len && (hdr_len == len)) {
unsigned int pull_size;
- pull_size = min((unsigned int)4, skb->data_len);
+ pull_size = min_t(unsigned int, 4, skb->data_len);
if (!__pskb_pull_tail(skb, pull_size)) {
e_err("__pskb_pull_tail failed.\n");
dev_kfree_skb_any(skb);
* need: count + 2 desc gap to keep tail from touching
* head, otherwise try next time
*/
- if (e1000_maybe_stop_tx(netdev, count + 2))
+ if (e1000_maybe_stop_tx(tx_ring, count + 2))
return NETDEV_TX_BUSY;
if (vlan_tx_tag_present(skb)) {
first = tx_ring->next_to_use;
- tso = e1000_tso(adapter, skb);
+ tso = e1000_tso(tx_ring, skb);
if (tso < 0) {
dev_kfree_skb_any(skb);
return NETDEV_TX_OK;
if (tso)
tx_flags |= E1000_TX_FLAGS_TSO;
- else if (e1000_tx_csum(adapter, skb))
+ else if (e1000_tx_csum(tx_ring, skb))
tx_flags |= E1000_TX_FLAGS_CSUM;
/*
if (skb->protocol == htons(ETH_P_IP))
tx_flags |= E1000_TX_FLAGS_IPV4;
+ if (unlikely(skb->no_fcs))
+ tx_flags |= E1000_TX_FLAGS_NO_FCS;
+
/* if count is 0 then mapping error has occurred */
- count = e1000_tx_map(adapter, skb, first, max_per_txd, nr_frags, mss);
+ count = e1000_tx_map(tx_ring, skb, first, max_per_txd, nr_frags, mss);
if (count) {
netdev_sent_queue(netdev, skb->len);
- e1000_tx_queue(adapter, tx_flags, count);
+ e1000_tx_queue(tx_ring, tx_flags, count);
/* Make sure there is space in the ring for the next send. */
- e1000_maybe_stop_tx(netdev, MAX_SKB_FRAGS + 2);
+ e1000_maybe_stop_tx(tx_ring, MAX_SKB_FRAGS + 2);
} else {
dev_kfree_skb_any(skb);
int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;
/* Jumbo frame support */
- if ((max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) &&
- !(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
- e_err("Jumbo Frames not supported.\n");
- return -EINVAL;
+ if (max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) {
+ if (!(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
+ e_err("Jumbo Frames not supported.\n");
+ return -EINVAL;
+ }
+
+ /*
+ * IP payload checksum (enabled with jumbos/packet-split when
+ * Rx checksum is enabled) and generation of RSS hash is
+ * mutually exclusive in the hardware.
+ */
+ if ((netdev->features & NETIF_F_RXCSUM) &&
+ (netdev->features & NETIF_F_RXHASH)) {
+ e_err("Jumbo frames cannot be enabled when both receive checksum offload and receive hashing are enabled. Disable one of the receive offload features before enabling jumbos.\n");
+ return -EINVAL;
+ }
}
/* Supported frame sizes */
/* Enable access to wakeup registers on and set page to BM_WUC_PAGE */
retval = e1000_enable_phy_wakeup_reg_access_bm(hw, &wuc_enable);
if (retval)
- goto out;
+ goto release;
/* copy MAC MTA to PHY MTA - only needed for pchlan */
for (i = 0; i < adapter->hw.mac.mta_reg_count; i++) {
retval = e1000_disable_phy_wakeup_reg_access_bm(hw, &wuc_enable);
if (retval)
e_err("Could not set PHY Host Wakeup bit\n");
-out:
+release:
hw->phy.ops.release(hw);
return retval;
ret_val = e1000_read_pba_string_generic(hw, pba_str,
E1000_PBANUM_LENGTH);
if (ret_val)
- strncpy((char *)pba_str, "Unknown", sizeof(pba_str) - 1);
+ strlcpy((char *)pba_str, "Unknown", sizeof(pba_str));
e_info("MAC: %d, PHY: %d, PBA No: %s\n",
hw->mac.type, hw->phy.type, pba_str);
}
return;
ret_val = e1000_read_nvm(hw, NVM_INIT_CONTROL2_REG, 1, &buf);
- if (!ret_val && (!(le16_to_cpu(buf) & (1 << 0)))) {
+ le16_to_cpus(&buf);
+ if (!ret_val && (!(buf & (1 << 0)))) {
/* Deep Smart Power Down (DSPD) */
dev_warn(&adapter->pdev->dev,
"Warning: detected DSPD enabled in EEPROM\n");
}
static int e1000_set_features(struct net_device *netdev,
- netdev_features_t features)
+ netdev_features_t features)
{
struct e1000_adapter *adapter = netdev_priv(netdev);
netdev_features_t changed = features ^ netdev->features;
adapter->flags |= FLAG_TSO_FORCE;
if (!(changed & (NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_TX |
- NETIF_F_RXCSUM)))
+ NETIF_F_RXCSUM | NETIF_F_RXHASH | NETIF_F_RXFCS |
+ NETIF_F_RXALL)))
return 0;
+ /*
+ * IP payload checksum (enabled with jumbos/packet-split when Rx
+ * checksum is enabled) and generation of RSS hash is mutually
+ * exclusive in the hardware.
+ */
+ if (adapter->rx_ps_pages &&
+ (features & NETIF_F_RXCSUM) && (features & NETIF_F_RXHASH)) {
+ e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames. Disable jumbos or enable only one of the receive offload features.\n");
+ return -EINVAL;
+ }
+
+ if (changed & NETIF_F_RXFCS) {
+ if (features & NETIF_F_RXFCS) {
+ adapter->flags2 &= ~FLAG2_CRC_STRIPPING;
+ } else {
+ /* We need to take it back to defaults, which might mean
+ * stripping is still disabled at the adapter level.
+ */
+ if (adapter->flags2 & FLAG2_DFLT_CRC_STRIPPING)
+ adapter->flags2 |= FLAG2_CRC_STRIPPING;
+ else
+ adapter->flags2 &= ~FLAG2_CRC_STRIPPING;
+ }
+ }
+
+ netdev->features = features;
+
if (netif_running(netdev))
e1000e_reinit_locked(adapter);
else
const struct e1000_info *ei = e1000_info_tbl[ent->driver_data];
resource_size_t mmio_start, mmio_len;
resource_size_t flash_start, flash_len;
-
static int cards_found;
u16 aspm_disable_flag = 0;
int i, err, pci_using_dac;
e1000e_set_ethtool_ops(netdev);
netdev->watchdog_timeo = 5 * HZ;
netif_napi_add(netdev, &adapter->napi, e1000_clean, 64);
- strncpy(netdev->name, pci_name(pdev), sizeof(netdev->name) - 1);
+ strlcpy(netdev->name, pci_name(pdev), sizeof(netdev->name));
netdev->mem_start = mmio_start;
netdev->mem_end = mmio_start + mmio_len;
adapter->hw.phy.ms_type = e1000_ms_hw_default;
}
- if (e1000_check_reset_block(&adapter->hw))
+ if (hw->phy.ops.check_reset_block(hw))
e_info("PHY reset is blocked due to SOL/IDER session.\n");
/* Set initial default active device features */
NETIF_F_HW_VLAN_TX |
NETIF_F_TSO |
NETIF_F_TSO6 |
+ NETIF_F_RXHASH |
NETIF_F_RXCSUM |
NETIF_F_HW_CSUM);
/* Set user-changeable features (subset of all device features) */
netdev->hw_features = netdev->features;
+ netdev->hw_features |= NETIF_F_RXFCS;
+ netdev->priv_flags |= IFF_SUPP_NOFCS;
+ netdev->hw_features |= NETIF_F_RXALL;
if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER)
netdev->features |= NETIF_F_HW_VLAN_FILTER;
} else if (adapter->flags & FLAG_APME_IN_CTRL3) {
if (adapter->flags & FLAG_APME_CHECK_PORT_B &&
(adapter->hw.bus.func == 1))
- e1000_read_nvm(&adapter->hw,
- NVM_INIT_CONTROL3_PORT_B, 1, &eeprom_data);
+ e1000_read_nvm(&adapter->hw, NVM_INIT_CONTROL3_PORT_B,
+ 1, &eeprom_data);
else
- e1000_read_nvm(&adapter->hw,
- NVM_INIT_CONTROL3_PORT_A, 1, &eeprom_data);
+ e1000_read_nvm(&adapter->hw, NVM_INIT_CONTROL3_PORT_A,
+ 1, &eeprom_data);
}
/* fetch WoL from EEPROM */
if (!(adapter->flags & FLAG_HAS_AMT))
e1000e_get_hw_control(adapter);
- strncpy(netdev->name, "eth%d", sizeof(netdev->name) - 1);
+ strlcpy(netdev->name, "eth%d", sizeof(netdev->name));
err = register_netdev(netdev);
if (err)
goto err_register;
if (!(adapter->flags & FLAG_HAS_AMT))
e1000e_release_hw_control(adapter);
err_eeprom:
- if (!e1000_check_reset_block(&adapter->hw))
+ if (!hw->phy.ops.check_reset_block(hw))
e1000_phy_hw_reset(&adapter->hw);
err_hw_init:
kfree(adapter->tx_ring);
{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH2_LV_LM), board_pch2lan },
{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH2_LV_V), board_pch2lan },
- { } /* terminate list */
+ { 0, 0, 0, 0, 0, 0, 0 } /* terminate list */
};
MODULE_DEVICE_TABLE(pci, e1000_pci_tbl);
.probe = e1000_probe,
.remove = __devexit_p(e1000_remove),
#ifdef CONFIG_PM
- .driver.pm = &e1000_pm_ops,
+ .driver = {
+ .pm = &e1000_pm_ops,
+ },
#endif
.shutdown = e1000_shutdown,
.err_handler = &e1000_err_handler
int ret;
pr_info("Intel(R) PRO/1000 Network Driver - %s\n",
e1000e_driver_version);
- pr_info("Copyright(c) 1999 - 2011 Intel Corporation.\n");
+ pr_info("Copyright(c) 1999 - 2012 Intel Corporation.\n");
ret = pci_register_driver(&e1000_driver);
return ret;
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);
-/* e1000_main.c */
+/* netdev.c */
#include <asm/byteorder.h>
#include <asm/uaccess.h>
- #define cas_page_map(x) kmap_atomic((x), KM_SKB_DATA_SOFTIRQ)
- #define cas_page_unmap(x) kunmap_atomic((x), KM_SKB_DATA_SOFTIRQ)
+ #define cas_page_map(x) kmap_atomic((x))
+ #define cas_page_unmap(x) kunmap_atomic((x))
#define CAS_NCPUS num_online_cpus()
#define cas_skb_release(x) netif_rx(x)
cp->fw_data = vmalloc(cp->fw_size);
if (!cp->fw_data) {
err = -ENOMEM;
- pr_err("\"%s\" Failed %d\n", fw_name, err);
goto out;
}
memcpy(cp->fw_data, &fw->data[2], cp->fw_size);
else
alloclen = max(hlen, RX_COPY_MIN);
- skb = dev_alloc_skb(alloclen + swivel + cp->crc_size);
+ skb = netdev_alloc_skb(cp->dev, alloclen + swivel + cp->crc_size);
if (skb == NULL)
return -1;
dev = alloc_etherdev(sizeof(*cp));
if (!dev) {
- dev_err(&pdev->dev, "Etherdev alloc failed, aborting\n");
err = -ENOMEM;
goto err_out_disable_pdev;
}
/* crc offload */
if (likely(lport->crc_offload)) {
- skb->ip_summed = CHECKSUM_PARTIAL;
+ skb->ip_summed = CHECKSUM_UNNECESSARY;
skb->csum_start = skb_headroom(skb);
skb->csum_offset = skb->len;
crc = 0;
return -ENOMEM;
}
frag = &skb_shinfo(skb)->frags[skb_shinfo(skb)->nr_frags - 1];
- cp = kmap_atomic(skb_frag_page(frag), KM_SKB_DATA_SOFTIRQ)
+ cp = kmap_atomic(skb_frag_page(frag))
+ frag->page_offset;
} else {
cp = (struct fcoe_crc_eof *)skb_put(skb, tlen);
cp->fcoe_crc32 = cpu_to_le32(~crc);
if (skb_is_nonlinear(skb)) {
- kunmap_atomic(cp, KM_SKB_DATA_SOFTIRQ);
+ kunmap_atomic(cp);
cp = NULL;
}
#include <scsi/scsi_devinfo.h>
#include <scsi/scsi_dbg.h>
+/*
+ * All wire protocol details (storage protocol between the guest and the host)
+ * are consolidated here.
+ *
+ * Begin protocol definitions.
+ */
-#define STORVSC_MIN_BUF_NR 64
-#define STORVSC_RING_BUFFER_SIZE (20*PAGE_SIZE)
-static int storvsc_ringbuffer_size = STORVSC_RING_BUFFER_SIZE;
-
-module_param(storvsc_ringbuffer_size, int, S_IRUGO);
-MODULE_PARM_DESC(storvsc_ringbuffer_size, "Ring buffer size (bytes)");
-
-/* to alert the user that structure sizes may be mismatched even though the */
-/* protocol versions match. */
-
-
-#define REVISION_STRING(REVISION_) #REVISION_
-#define FILL_VMSTOR_REVISION(RESULT_LVALUE_) \
- do { \
- char *revision_string \
- = REVISION_STRING($Rev : 6 $) + 6; \
- RESULT_LVALUE_ = 0; \
- while (*revision_string >= '0' \
- && *revision_string <= '9') { \
- RESULT_LVALUE_ *= 10; \
- RESULT_LVALUE_ += *revision_string - '0'; \
- revision_string++; \
- } \
- } while (0)
-
-/* Major/minor macros. Minor version is in LSB, meaning that earlier flat */
-/* version numbers will be interpreted as "0.x" (i.e., 1 becomes 0.1). */
-#define VMSTOR_PROTOCOL_MAJOR(VERSION_) (((VERSION_) >> 8) & 0xff)
-#define VMSTOR_PROTOCOL_MINOR(VERSION_) (((VERSION_)) & 0xff)
-#define VMSTOR_PROTOCOL_VERSION(MAJOR_, MINOR_) ((((MAJOR_) & 0xff) << 8) | \
- (((MINOR_) & 0xff)))
-#define VMSTOR_INVALID_PROTOCOL_VERSION (-1)
-
-/* Version history: */
-/* V1 Beta 0.1 */
-/* V1 RC < 2008/1/31 1.0 */
-/* V1 RC > 2008/1/31 2.0 */
-#define VMSTOR_PROTOCOL_VERSION_CURRENT VMSTOR_PROTOCOL_VERSION(4, 2)
-
-
-
+/*
+ * Version history:
+ * V1 Beta: 0.1
+ * V1 RC < 2008/1/31: 1.0
+ * V1 RC > 2008/1/31: 2.0
+ * Win7: 4.2
+ */
-/* This will get replaced with the max transfer length that is possible on */
-/* the host adapter. */
-/* The max transfer length will be published when we offer a vmbus channel. */
-#define MAX_TRANSFER_LENGTH 0x40000
-#define DEFAULT_PACKET_SIZE (sizeof(struct vmdata_gpa_direct) + \
- sizeof(struct vstor_packet) + \
- sizesizeof(u64) * (MAX_TRANSFER_LENGTH / PAGE_SIZE)))
+#define VMSTOR_CURRENT_MAJOR 4
+#define VMSTOR_CURRENT_MINOR 2
/* Packet structure describing virtual storage requests. */
* this remains the same across the write regardless of 32/64 bit
* note: it's patterned off the SCSI_PASS_THROUGH structure
*/
-#define CDB16GENERIC_LENGTH 0x10
-
-#ifndef SENSE_BUFFER_SIZE
-#define SENSE_BUFFER_SIZE 0x12
-#endif
-
-#define MAX_DATA_BUF_LEN_WITH_PADDING 0x14
+#define STORVSC_MAX_CMD_LEN 0x10
+#define STORVSC_SENSE_BUFFER_SIZE 0x12
+#define STORVSC_MAX_BUF_LEN_WITH_PADDING 0x14
struct vmscsi_request {
- unsigned short length;
- unsigned char srb_status;
- unsigned char scsi_status;
+ u16 length;
+ u8 srb_status;
+ u8 scsi_status;
- unsigned char port_number;
- unsigned char path_id;
- unsigned char target_id;
- unsigned char lun;
+ u8 port_number;
+ u8 path_id;
+ u8 target_id;
+ u8 lun;
- unsigned char cdb_length;
- unsigned char sense_info_length;
- unsigned char data_in;
- unsigned char reserved;
+ u8 cdb_length;
+ u8 sense_info_length;
+ u8 data_in;
+ u8 reserved;
- unsigned int data_transfer_length;
+ u32 data_transfer_length;
union {
- unsigned char cdb[CDB16GENERIC_LENGTH];
- unsigned char sense_data[SENSE_BUFFER_SIZE];
- unsigned char reserved_array[MAX_DATA_BUF_LEN_WITH_PADDING];
+ u8 cdb[STORVSC_MAX_CMD_LEN];
+ u8 sense_data[STORVSC_SENSE_BUFFER_SIZE];
+ u8 reserved_array[STORVSC_MAX_BUF_LEN_WITH_PADDING];
};
} __attribute((packed));
* properties of the channel.
*/
struct vmstorage_channel_properties {
- unsigned short protocol_version;
- unsigned char path_id;
- unsigned char target_id;
+ u16 protocol_version;
+ u8 path_id;
+ u8 target_id;
/* Note: port number is only really known on the client side */
- unsigned int port_number;
- unsigned int flags;
- unsigned int max_transfer_bytes;
+ u32 port_number;
+ u32 flags;
+ u32 max_transfer_bytes;
- /* This id is unique for each channel and will correspond with */
- /* vendor specific data in the inquirydata */
- unsigned long long unique_id;
+ /*
+ * This id is unique for each channel and will correspond with
+ * vendor specific data in the inquiry data.
+ */
+
+ u64 unique_id;
} __packed;
/* This structure is sent during the storage protocol negotiations. */
struct vmstorage_protocol_version {
/* Major (MSW) and minor (LSW) version numbers. */
- unsigned short major_minor;
+ u16 major_minor;
/*
* Revision number is auto-incremented whenever this file is changed
* (See FILL_VMSTOR_REVISION macro above). Mismatch does not
* definitely indicate incompatibility--but it does indicate mismatched
* builds.
+ * This is only used on the windows side. Just set it to 0.
*/
- unsigned short revision;
+ u16 revision;
} __packed;
/* Channel Property Flags */
enum vstor_packet_operation operation;
/* Flags - see below for values */
- unsigned int flags;
+ u32 flags;
/* Status of the request returned from the server side. */
- unsigned int status;
+ u32 status;
/* Data payload area */
union {
};
} __packed;
-/* Packet flags */
/*
+ * Packet Flags:
+ *
* This flag indicates that the server should send back a completion for this
* packet.
*/
+
#define REQUEST_COMPLETION_FLAG 0x1
-/* This is the set of flags that the vsc can set in any packets it sends */
-#define VSC_LEGAL_FLAGS (REQUEST_COMPLETION_FLAG)
+/* Matches Windows-end */
+enum storvsc_request_type {
+ WRITE_TYPE = 0,
+ READ_TYPE,
+ UNKNOWN_TYPE,
+};
+
+/*
+ * SRB status codes and masks; a subset of the codes used here.
+ */
+
+#define SRB_STATUS_AUTOSENSE_VALID 0x80
+#define SRB_STATUS_INVALID_LUN 0x20
+#define SRB_STATUS_SUCCESS 0x01
+#define SRB_STATUS_ERROR 0x04
+
+/*
+ * This is the end of Protocol specific defines.
+ */
+
+
+/*
+ * We setup a mempool to allocate request structures for this driver
+ * on a per-lun basis. The following define specifies the number of
+ * elements in the pool.
+ */
+#define STORVSC_MIN_BUF_NR 64
+static int storvsc_ringbuffer_size = (20 * PAGE_SIZE);
-/* Defines */
+module_param(storvsc_ringbuffer_size, int, S_IRUGO);
+MODULE_PARM_DESC(storvsc_ringbuffer_size, "Ring buffer size (bytes)");
#define STORVSC_MAX_IO_REQUESTS 128
#define STORVSC_MAX_LUNS_PER_TARGET 64
#define STORVSC_MAX_TARGETS 1
#define STORVSC_MAX_CHANNELS 1
-#define STORVSC_MAX_CMD_LEN 16
-/* Matches Windows-end */
-enum storvsc_request_type {
- WRITE_TYPE,
- READ_TYPE,
- UNKNOWN_TYPE,
-};
-struct hv_storvsc_request {
+struct storvsc_cmd_request {
+ struct list_head entry;
+ struct scsi_cmnd *cmd;
+
+ unsigned int bounce_sgl_count;
+ struct scatterlist *bounce_sgl;
+
struct hv_device *device;
/* Synchronize the request/response if needed */
struct completion wait_event;
unsigned char *sense_buffer;
- void *context;
- void (*on_io_completion)(struct hv_storvsc_request *request);
struct hv_multipage_buffer data_buffer;
-
struct vstor_packet vstor_packet;
};
unsigned char target_id;
/* Used for vsc/vsp channel reset process */
- struct hv_storvsc_request init_request;
- struct hv_storvsc_request reset_request;
+ struct storvsc_cmd_request init_request;
+ struct storvsc_cmd_request reset_request;
};
struct stor_mem_pools {
unsigned char target;
};
-struct storvsc_cmd_request {
- struct list_head entry;
- struct scsi_cmnd *cmd;
-
- unsigned int bounce_sgl_count;
- struct scatterlist *bounce_sgl;
-
- struct hv_storvsc_request request;
-};
-
struct storvsc_scan_work {
struct work_struct work;
struct Scsi_Host *host;
kfree(wrk);
}
+/*
+ * Major/minor macros. Minor version is in LSB, meaning that earlier flat
+ * version numbers will be interpreted as "0.x" (i.e., 1 becomes 0.1).
+ */
+
+static inline u16 storvsc_get_version(u8 major, u8 minor)
+{
+ u16 version;
+
+ version = ((major << 8) | minor);
+ return version;
+}
+
+/*
+ * We can get incoming messages from the host that are not in response to
+ * messages that we have sent out. An example of this would be messages
+ * received by the guest to notify dynamic addition/removal of LUNs. To
+ * deal with potential race conditions where the driver may be in the
+ * midst of being unloaded when we might receive an unsolicited message
+ * from the host, we have implemented a mechanism to gurantee sequential
+ * consistency:
+ *
+ * 1) Once the device is marked as being destroyed, we will fail all
+ * outgoing messages.
+ * 2) We permit incoming messages when the device is being destroyed,
+ * only to properly account for messages already sent out.
+ */
+
static inline struct storvsc_device *get_out_stor_device(
struct hv_device *device)
{
}
-static int storvsc_channel_init(struct hv_device *device)
+static void destroy_bounce_buffer(struct scatterlist *sgl,
+ unsigned int sg_count)
{
- struct storvsc_device *stor_device;
- struct hv_storvsc_request *request;
- struct vstor_packet *vstor_packet;
- int ret, t;
+ int i;
+ struct page *page_buf;
- stor_device = get_out_stor_device(device);
- if (!stor_device)
- return -ENODEV;
+ for (i = 0; i < sg_count; i++) {
+ page_buf = sg_page((&sgl[i]));
+ if (page_buf != NULL)
+ __free_page(page_buf);
+ }
- request = &stor_device->init_request;
- vstor_packet = &request->vstor_packet;
+ kfree(sgl);
+}
- /*
- * Now, initiate the vsc/vsp initialization protocol on the open
- * channel
- */
- memset(request, 0, sizeof(struct hv_storvsc_request));
- init_completion(&request->wait_event);
- vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
- vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+static int do_bounce_buffer(struct scatterlist *sgl, unsigned int sg_count)
+{
+ int i;
- ret = vmbus_sendpacket(device->channel, vstor_packet,
- sizeof(struct vstor_packet),
- (unsigned long)request,
- VM_PKT_DATA_INBAND,
- VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
- if (ret != 0)
- goto cleanup;
+ /* No need to check */
+ if (sg_count < 2)
+ return -1;
- t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
- if (t == 0) {
- ret = -ETIMEDOUT;
- goto cleanup;
+ /* We have at least 2 sg entries */
+ for (i = 0; i < sg_count; i++) {
+ if (i == 0) {
+ /* make sure 1st one does not have hole */
+ if (sgl[i].offset + sgl[i].length != PAGE_SIZE)
+ return i;
+ } else if (i == sg_count - 1) {
+ /* make sure last one does not have hole */
+ if (sgl[i].offset != 0)
+ return i;
+ } else {
+ /* make sure no hole in the middle */
+ if (sgl[i].length != PAGE_SIZE || sgl[i].offset != 0)
+ return i;
+ }
}
+ return -1;
+}
- if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
- vstor_packet->status != 0)
- goto cleanup;
-
-
- /* reuse the packet for version range supported */
- memset(vstor_packet, 0, sizeof(struct vstor_packet));
- vstor_packet->operation = VSTOR_OPERATION_QUERY_PROTOCOL_VERSION;
- vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+static struct scatterlist *create_bounce_buffer(struct scatterlist *sgl,
+ unsigned int sg_count,
+ unsigned int len,
+ int write)
+{
+ int i;
+ int num_pages;
+ struct scatterlist *bounce_sgl;
+ struct page *page_buf;
+ unsigned int buf_len = ((write == WRITE_TYPE) ? 0 : PAGE_SIZE);
- vstor_packet->version.major_minor = VMSTOR_PROTOCOL_VERSION_CURRENT;
- FILL_VMSTOR_REVISION(vstor_packet->version.revision);
+ num_pages = ALIGN(len, PAGE_SIZE) >> PAGE_SHIFT;
- ret = vmbus_sendpacket(device->channel, vstor_packet,
- sizeof(struct vstor_packet),
- (unsigned long)request,
- VM_PKT_DATA_INBAND,
- VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
- if (ret != 0)
- goto cleanup;
+ bounce_sgl = kcalloc(num_pages, sizeof(struct scatterlist), GFP_ATOMIC);
+ if (!bounce_sgl)
+ return NULL;
- t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
- if (t == 0) {
- ret = -ETIMEDOUT;
- goto cleanup;
+ for (i = 0; i < num_pages; i++) {
+ page_buf = alloc_page(GFP_ATOMIC);
+ if (!page_buf)
+ goto cleanup;
+ sg_set_page(&bounce_sgl[i], page_buf, buf_len, 0);
}
- if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
- vstor_packet->status != 0)
- goto cleanup;
-
+ return bounce_sgl;
- memset(vstor_packet, 0, sizeof(struct vstor_packet));
- vstor_packet->operation = VSTOR_OPERATION_QUERY_PROPERTIES;
- vstor_packet->flags = REQUEST_COMPLETION_FLAG;
- vstor_packet->storage_channel_properties.port_number =
- stor_device->port_number;
+cleanup:
+ destroy_bounce_buffer(bounce_sgl, num_pages);
+ return NULL;
+}
- ret = vmbus_sendpacket(device->channel, vstor_packet,
- sizeof(struct vstor_packet),
- (unsigned long)request,
- VM_PKT_DATA_INBAND,
- VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
++/* Disgusting wrapper functions */
++static inline unsigned long sg_kmap_atomic(struct scatterlist *sgl, int idx)
++{
++ void *addr = kmap_atomic(sg_page(sgl + idx));
++ return (unsigned long)addr;
++}
+
- if (ret != 0)
- goto cleanup;
++static inline void sg_kunmap_atomic(unsigned long addr)
++{
++ kunmap_atomic((void *)addr);
++}
+
- t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
- if (t == 0) {
- ret = -ETIMEDOUT;
- goto cleanup;
- }
+
- if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
- vstor_packet->status != 0)
- goto cleanup;
+/* Assume the original sgl has enough room */
+static unsigned int copy_from_bounce_buffer(struct scatterlist *orig_sgl,
+ struct scatterlist *bounce_sgl,
+ unsigned int orig_sgl_count,
+ unsigned int bounce_sgl_count)
+{
+ int i;
+ int j = 0;
+ unsigned long src, dest;
+ unsigned int srclen, destlen, copylen;
+ unsigned int total_copied = 0;
+ unsigned long bounce_addr = 0;
+ unsigned long dest_addr = 0;
+ unsigned long flags;
- stor_device->path_id = vstor_packet->storage_channel_properties.path_id;
- stor_device->target_id
- = vstor_packet->storage_channel_properties.target_id;
+ local_irq_save(flags);
- memset(vstor_packet, 0, sizeof(struct vstor_packet));
- vstor_packet->operation = VSTOR_OPERATION_END_INITIALIZATION;
- vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+ for (i = 0; i < orig_sgl_count; i++) {
- dest_addr = (unsigned long)kmap_atomic(sg_page((&orig_sgl[i])),
- KM_IRQ0) + orig_sgl[i].offset;
++ dest_addr = sg_kmap_atomic(orig_sgl,i) + orig_sgl[i].offset;
+ dest = dest_addr;
+ destlen = orig_sgl[i].length;
- ret = vmbus_sendpacket(device->channel, vstor_packet,
- sizeof(struct vstor_packet),
- (unsigned long)request,
- VM_PKT_DATA_INBAND,
- VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
+ if (bounce_addr == 0)
- bounce_addr =
- (unsigned long)kmap_atomic(sg_page((&bounce_sgl[j])),
- KM_IRQ0);
++ bounce_addr = sg_kmap_atomic(bounce_sgl,j);
- if (ret != 0)
- goto cleanup;
+ while (destlen) {
+ src = bounce_addr + bounce_sgl[j].offset;
+ srclen = bounce_sgl[j].length - bounce_sgl[j].offset;
- t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
- if (t == 0) {
- ret = -ETIMEDOUT;
- goto cleanup;
- }
+ copylen = min(srclen, destlen);
+ memcpy((void *)dest, (void *)src, copylen);
+
+ total_copied += copylen;
+ bounce_sgl[j].offset += copylen;
+ destlen -= copylen;
+ dest += copylen;
+
+ if (bounce_sgl[j].offset == bounce_sgl[j].length) {
+ /* full */
- kunmap_atomic((void *)bounce_addr, KM_IRQ0);
++ sg_kunmap_atomic(bounce_addr);
+ j++;
+
+ /*
+ * It is possible that the number of elements
+ * in the bounce buffer may not be equal to
+ * the number of elements in the original
+ * scatter list. Handle this correctly.
+ */
+
+ if (j == bounce_sgl_count) {
+ /*
+ * We are done; cleanup and return.
+ */
- kunmap_atomic((void *)(dest_addr -
- orig_sgl[i].offset),
- KM_IRQ0);
++ sg_kunmap_atomic(dest_addr - orig_sgl[i].offset);
+ local_irq_restore(flags);
+ return total_copied;
+ }
+
+ /* if we need to use another bounce buffer */
+ if (destlen || i != orig_sgl_count - 1)
- bounce_addr =
- (unsigned long)kmap_atomic(
- sg_page((&bounce_sgl[j])), KM_IRQ0);
++ bounce_addr = sg_kmap_atomic(bounce_sgl,j);
+ } else if (destlen == 0 && i == orig_sgl_count - 1) {
+ /* unmap the last bounce that is < PAGE_SIZE */
- kunmap_atomic((void *)bounce_addr, KM_IRQ0);
++ sg_kunmap_atomic(bounce_addr);
+ }
+ }
+
- kunmap_atomic((void *)(dest_addr - orig_sgl[i].offset),
- KM_IRQ0);
++ sg_kunmap_atomic(dest_addr - orig_sgl[i].offset);
+ }
+
+ local_irq_restore(flags);
+
+ return total_copied;
+}
+
+/* Assume the bounce_sgl has enough room ie using the create_bounce_buffer() */
+static unsigned int copy_to_bounce_buffer(struct scatterlist *orig_sgl,
+ struct scatterlist *bounce_sgl,
+ unsigned int orig_sgl_count)
+{
+ int i;
+ int j = 0;
+ unsigned long src, dest;
+ unsigned int srclen, destlen, copylen;
+ unsigned int total_copied = 0;
+ unsigned long bounce_addr = 0;
+ unsigned long src_addr = 0;
+ unsigned long flags;
+
+ local_irq_save(flags);
+
+ for (i = 0; i < orig_sgl_count; i++) {
- src_addr = (unsigned long)kmap_atomic(sg_page((&orig_sgl[i])),
- KM_IRQ0) + orig_sgl[i].offset;
++ src_addr = sg_kmap_atomic(orig_sgl,i) + orig_sgl[i].offset;
+ src = src_addr;
+ srclen = orig_sgl[i].length;
+
+ if (bounce_addr == 0)
- bounce_addr =
- (unsigned long)kmap_atomic(sg_page((&bounce_sgl[j])),
- KM_IRQ0);
++ bounce_addr = sg_kmap_atomic(bounce_sgl,j);
+
+ while (srclen) {
+ /* assume bounce offset always == 0 */
+ dest = bounce_addr + bounce_sgl[j].length;
+ destlen = PAGE_SIZE - bounce_sgl[j].length;
+
+ copylen = min(srclen, destlen);
+ memcpy((void *)dest, (void *)src, copylen);
+
+ total_copied += copylen;
+ bounce_sgl[j].length += copylen;
+ srclen -= copylen;
+ src += copylen;
+
+ if (bounce_sgl[j].length == PAGE_SIZE) {
+ /* full..move to next entry */
- kunmap_atomic((void *)bounce_addr, KM_IRQ0);
++ sg_kunmap_atomic(bounce_addr);
+ j++;
+
+ /* if we need to use another bounce buffer */
+ if (srclen || i != orig_sgl_count - 1)
- bounce_addr =
- (unsigned long)kmap_atomic(
- sg_page((&bounce_sgl[j])), KM_IRQ0);
++ bounce_addr = sg_kmap_atomic(bounce_sgl,j);
+
+ } else if (srclen == 0 && i == orig_sgl_count - 1) {
+ /* unmap the last bounce that is < PAGE_SIZE */
- kunmap_atomic((void *)bounce_addr, KM_IRQ0);
++ sg_kunmap_atomic(bounce_addr);
+ }
+ }
+
- kunmap_atomic((void *)(src_addr - orig_sgl[i].offset), KM_IRQ0);
++ sg_kunmap_atomic(src_addr - orig_sgl[i].offset);
+ }
+
+ local_irq_restore(flags);
+
+ return total_copied;
+}
+
+static int storvsc_channel_init(struct hv_device *device)
+{
+ struct storvsc_device *stor_device;
+ struct storvsc_cmd_request *request;
+ struct vstor_packet *vstor_packet;
+ int ret, t;
+
+ stor_device = get_out_stor_device(device);
+ if (!stor_device)
+ return -ENODEV;
+
+ request = &stor_device->init_request;
+ vstor_packet = &request->vstor_packet;
+
+ /*
+ * Now, initiate the vsc/vsp initialization protocol on the open
+ * channel
+ */
+ memset(request, 0, sizeof(struct storvsc_cmd_request));
+ init_completion(&request->wait_event);
+ vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
+ vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+
+ ret = vmbus_sendpacket(device->channel, vstor_packet,
+ sizeof(struct vstor_packet),
+ (unsigned long)request,
+ VM_PKT_DATA_INBAND,
+ VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
+ if (ret != 0)
+ goto cleanup;
+
+ t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
+ if (t == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
+
+ if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
+ vstor_packet->status != 0)
+ goto cleanup;
+
+
+ /* reuse the packet for version range supported */
+ memset(vstor_packet, 0, sizeof(struct vstor_packet));
+ vstor_packet->operation = VSTOR_OPERATION_QUERY_PROTOCOL_VERSION;
+ vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+
+ vstor_packet->version.major_minor =
+ storvsc_get_version(VMSTOR_CURRENT_MAJOR, VMSTOR_CURRENT_MINOR);
+
+ /*
+ * The revision number is only used in Windows; set it to 0.
+ */
+ vstor_packet->version.revision = 0;
+
+ ret = vmbus_sendpacket(device->channel, vstor_packet,
+ sizeof(struct vstor_packet),
+ (unsigned long)request,
+ VM_PKT_DATA_INBAND,
+ VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
+ if (ret != 0)
+ goto cleanup;
+
+ t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
+ if (t == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
+
+ if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
+ vstor_packet->status != 0)
+ goto cleanup;
+
+
+ memset(vstor_packet, 0, sizeof(struct vstor_packet));
+ vstor_packet->operation = VSTOR_OPERATION_QUERY_PROPERTIES;
+ vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+ vstor_packet->storage_channel_properties.port_number =
+ stor_device->port_number;
+
+ ret = vmbus_sendpacket(device->channel, vstor_packet,
+ sizeof(struct vstor_packet),
+ (unsigned long)request,
+ VM_PKT_DATA_INBAND,
+ VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
+
+ if (ret != 0)
+ goto cleanup;
+
+ t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
+ if (t == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
+
+ if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
+ vstor_packet->status != 0)
+ goto cleanup;
+
+ stor_device->path_id = vstor_packet->storage_channel_properties.path_id;
+ stor_device->target_id
+ = vstor_packet->storage_channel_properties.target_id;
+
+ memset(vstor_packet, 0, sizeof(struct vstor_packet));
+ vstor_packet->operation = VSTOR_OPERATION_END_INITIALIZATION;
+ vstor_packet->flags = REQUEST_COMPLETION_FLAG;
+
+ ret = vmbus_sendpacket(device->channel, vstor_packet,
+ sizeof(struct vstor_packet),
+ (unsigned long)request,
+ VM_PKT_DATA_INBAND,
+ VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
+
+ if (ret != 0)
+ goto cleanup;
+
+ t = wait_for_completion_timeout(&request->wait_event, 5*HZ);
+ if (t == 0) {
+ ret = -ETIMEDOUT;
+ goto cleanup;
+ }
if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
vstor_packet->status != 0)
return ret;
}
+
+static void storvsc_command_completion(struct storvsc_cmd_request *cmd_request)
+{
+ struct scsi_cmnd *scmnd = cmd_request->cmd;
+ struct hv_host_device *host_dev = shost_priv(scmnd->device->host);
+ void (*scsi_done_fn)(struct scsi_cmnd *);
+ struct scsi_sense_hdr sense_hdr;
+ struct vmscsi_request *vm_srb;
+ struct storvsc_scan_work *wrk;
+ struct stor_mem_pools *memp = scmnd->device->hostdata;
+
+ vm_srb = &cmd_request->vstor_packet.vm_srb;
+ if (cmd_request->bounce_sgl_count) {
+ if (vm_srb->data_in == READ_TYPE)
+ copy_from_bounce_buffer(scsi_sglist(scmnd),
+ cmd_request->bounce_sgl,
+ scsi_sg_count(scmnd),
+ cmd_request->bounce_sgl_count);
+ destroy_bounce_buffer(cmd_request->bounce_sgl,
+ cmd_request->bounce_sgl_count);
+ }
+
+ /*
+ * If there is an error; offline the device since all
+ * error recovery strategies would have already been
+ * deployed on the host side.
+ */
+ if (vm_srb->srb_status == SRB_STATUS_ERROR)
+ scmnd->result = DID_TARGET_FAILURE << 16;
+ else
+ scmnd->result = vm_srb->scsi_status;
+
+ /*
+ * If the LUN is invalid; remove the device.
+ */
+ if (vm_srb->srb_status == SRB_STATUS_INVALID_LUN) {
+ struct storvsc_device *stor_dev;
+ struct hv_device *dev = host_dev->dev;
+ struct Scsi_Host *host;
+
+ stor_dev = get_in_stor_device(dev);
+ host = stor_dev->host;
+
+ wrk = kmalloc(sizeof(struct storvsc_scan_work),
+ GFP_ATOMIC);
+ if (!wrk) {
+ scmnd->result = DID_TARGET_FAILURE << 16;
+ } else {
+ wrk->host = host;
+ wrk->lun = vm_srb->lun;
+ INIT_WORK(&wrk->work, storvsc_remove_lun);
+ schedule_work(&wrk->work);
+ }
+ }
+
+ if (scmnd->result) {
+ if (scsi_normalize_sense(scmnd->sense_buffer,
+ SCSI_SENSE_BUFFERSIZE, &sense_hdr))
+ scsi_print_sense_hdr("storvsc", &sense_hdr);
+ }
+
+ scsi_set_resid(scmnd,
+ cmd_request->data_buffer.len -
+ vm_srb->data_transfer_length);
+
+ scsi_done_fn = scmnd->scsi_done;
+
+ scmnd->host_scribble = NULL;
+ scmnd->scsi_done = NULL;
+
+ scsi_done_fn(scmnd);
+
+ mempool_free(cmd_request, memp->request_mempool);
+}
+
static void storvsc_on_io_completion(struct hv_device *device,
struct vstor_packet *vstor_packet,
- struct hv_storvsc_request *request)
+ struct storvsc_cmd_request *request)
{
struct storvsc_device *stor_device;
struct vstor_packet *stor_pkt;
*/
if ((stor_pkt->vm_srb.cdb[0] == INQUIRY) ||
- (stor_pkt->vm_srb.cdb[0] == MODE_SENSE)) {
+ (stor_pkt->vm_srb.cdb[0] == MODE_SENSE)) {
vstor_packet->vm_srb.scsi_status = 0;
- vstor_packet->vm_srb.srb_status = 0x1;
+ vstor_packet->vm_srb.srb_status = SRB_STATUS_SUCCESS;
}
vstor_packet->vm_srb.sense_info_length;
if (vstor_packet->vm_srb.scsi_status != 0 ||
- vstor_packet->vm_srb.srb_status != 1){
+ vstor_packet->vm_srb.srb_status != SRB_STATUS_SUCCESS){
dev_warn(&device->device,
"cmd 0x%x scsi status 0x%x srb status 0x%x\n",
stor_pkt->vm_srb.cdb[0],
if ((vstor_packet->vm_srb.scsi_status & 0xFF) == 0x02) {
/* CHECK_CONDITION */
- if (vstor_packet->vm_srb.srb_status & 0x80) {
+ if (vstor_packet->vm_srb.srb_status &
+ SRB_STATUS_AUTOSENSE_VALID) {
/* autosense data available */
dev_warn(&device->device,
"stor pkt %p autosense data valid - len %d\n",
stor_pkt->vm_srb.data_transfer_length =
vstor_packet->vm_srb.data_transfer_length;
- request->on_io_completion(request);
+ storvsc_command_completion(request);
if (atomic_dec_and_test(&stor_device->num_outstanding_req) &&
stor_device->drain_notify)
static void storvsc_on_receive(struct hv_device *device,
struct vstor_packet *vstor_packet,
- struct hv_storvsc_request *request)
+ struct storvsc_cmd_request *request)
{
struct storvsc_scan_work *work;
struct storvsc_device *stor_device;
u32 bytes_recvd;
u64 request_id;
unsigned char packet[ALIGN(sizeof(struct vstor_packet), 8)];
- struct hv_storvsc_request *request;
+ struct storvsc_cmd_request *request;
int ret;
&bytes_recvd, &request_id);
if (ret == 0 && bytes_recvd > 0) {
- request = (struct hv_storvsc_request *)
+ request = (struct storvsc_cmd_request *)
(unsigned long)request_id;
if ((request == &stor_device->init_request) ||
memset(&props, 0, sizeof(struct vmstorage_channel_properties));
- /* Open the channel */
ret = vmbus_open(device->channel,
ring_size,
ring_size,
}
static int storvsc_do_io(struct hv_device *device,
- struct hv_storvsc_request *request)
+ struct storvsc_cmd_request *request)
{
struct storvsc_device *stor_device;
struct vstor_packet *vstor_packet;
vstor_packet->vm_srb.length = sizeof(struct vmscsi_request);
- vstor_packet->vm_srb.sense_info_length = SENSE_BUFFER_SIZE;
-
-
- vstor_packet->vm_srb.data_transfer_length =
- request->data_buffer.len;
-
- vstor_packet->operation = VSTOR_OPERATION_EXECUTE_SRB;
-
- if (request->data_buffer.len) {
- ret = vmbus_sendpacket_multipagebuffer(device->channel,
- &request->data_buffer,
- vstor_packet,
- sizeof(struct vstor_packet),
- (unsigned long)request);
- } else {
- ret = vmbus_sendpacket(device->channel, vstor_packet,
- sizeof(struct vstor_packet),
- (unsigned long)request,
- VM_PKT_DATA_INBAND,
- VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
- }
-
- if (ret != 0)
- return ret;
-
- atomic_inc(&stor_device->num_outstanding_req);
-
- return ret;
-}
-
-static void storvsc_get_ide_info(struct hv_device *dev, int *target, int *path)
-{
- *target =
- dev->dev_instance.b[5] << 8 | dev->dev_instance.b[4];
-
- *path =
- dev->dev_instance.b[3] << 24 |
- dev->dev_instance.b[2] << 16 |
- dev->dev_instance.b[1] << 8 | dev->dev_instance.b[0];
-}
-
-
-static int storvsc_device_alloc(struct scsi_device *sdevice)
-{
- struct stor_mem_pools *memp;
- int number = STORVSC_MIN_BUF_NR;
-
- memp = kzalloc(sizeof(struct stor_mem_pools), GFP_KERNEL);
- if (!memp)
- return -ENOMEM;
-
- memp->request_pool =
- kmem_cache_create(dev_name(&sdevice->sdev_dev),
- sizeof(struct storvsc_cmd_request), 0,
- SLAB_HWCACHE_ALIGN, NULL);
-
- if (!memp->request_pool)
- goto err0;
-
- memp->request_mempool = mempool_create(number, mempool_alloc_slab,
- mempool_free_slab,
- memp->request_pool);
-
- if (!memp->request_mempool)
- goto err1;
-
- sdevice->hostdata = memp;
-
- return 0;
-
-err1:
- kmem_cache_destroy(memp->request_pool);
-
-err0:
- kfree(memp);
- return -ENOMEM;
-}
-
-static void storvsc_device_destroy(struct scsi_device *sdevice)
-{
- struct stor_mem_pools *memp = sdevice->hostdata;
-
- mempool_destroy(memp->request_mempool);
- kmem_cache_destroy(memp->request_pool);
- kfree(memp);
- sdevice->hostdata = NULL;
-}
-
-static int storvsc_device_configure(struct scsi_device *sdevice)
-{
- scsi_adjust_queue_depth(sdevice, MSG_SIMPLE_TAG,
- STORVSC_MAX_IO_REQUESTS);
-
- blk_queue_max_segment_size(sdevice->request_queue, PAGE_SIZE);
-
- blk_queue_bounce_limit(sdevice->request_queue, BLK_BOUNCE_ANY);
-
- return 0;
-}
-
-static void destroy_bounce_buffer(struct scatterlist *sgl,
- unsigned int sg_count)
-{
- int i;
- struct page *page_buf;
-
- for (i = 0; i < sg_count; i++) {
- page_buf = sg_page((&sgl[i]));
- if (page_buf != NULL)
- __free_page(page_buf);
- }
-
- kfree(sgl);
-}
-
-static int do_bounce_buffer(struct scatterlist *sgl, unsigned int sg_count)
-{
- int i;
-
- /* No need to check */
- if (sg_count < 2)
- return -1;
-
- /* We have at least 2 sg entries */
- for (i = 0; i < sg_count; i++) {
- if (i == 0) {
- /* make sure 1st one does not have hole */
- if (sgl[i].offset + sgl[i].length != PAGE_SIZE)
- return i;
- } else if (i == sg_count - 1) {
- /* make sure last one does not have hole */
- if (sgl[i].offset != 0)
- return i;
- } else {
- /* make sure no hole in the middle */
- if (sgl[i].length != PAGE_SIZE || sgl[i].offset != 0)
- return i;
- }
- }
- return -1;
-}
-
-static struct scatterlist *create_bounce_buffer(struct scatterlist *sgl,
- unsigned int sg_count,
- unsigned int len,
- int write)
-{
- int i;
- int num_pages;
- struct scatterlist *bounce_sgl;
- struct page *page_buf;
- unsigned int buf_len = ((write == WRITE_TYPE) ? 0 : PAGE_SIZE);
-
- num_pages = ALIGN(len, PAGE_SIZE) >> PAGE_SHIFT;
-
- bounce_sgl = kcalloc(num_pages, sizeof(struct scatterlist), GFP_ATOMIC);
- if (!bounce_sgl)
- return NULL;
-
- for (i = 0; i < num_pages; i++) {
- page_buf = alloc_page(GFP_ATOMIC);
- if (!page_buf)
- goto cleanup;
- sg_set_page(&bounce_sgl[i], page_buf, buf_len, 0);
- }
-
- return bounce_sgl;
-
-cleanup:
- destroy_bounce_buffer(bounce_sgl, num_pages);
- return NULL;
-}
-
-
-/* Assume the original sgl has enough room */
-static unsigned int copy_from_bounce_buffer(struct scatterlist *orig_sgl,
- struct scatterlist *bounce_sgl,
- unsigned int orig_sgl_count,
- unsigned int bounce_sgl_count)
-{
- int i;
- int j = 0;
- unsigned long src, dest;
- unsigned int srclen, destlen, copylen;
- unsigned int total_copied = 0;
- unsigned long bounce_addr = 0;
- unsigned long dest_addr = 0;
- unsigned long flags;
-
- local_irq_save(flags);
-
- for (i = 0; i < orig_sgl_count; i++) {
- dest_addr = (unsigned long)kmap_atomic(sg_page((&orig_sgl[i])))
- + orig_sgl[i].offset;
- dest = dest_addr;
- destlen = orig_sgl[i].length;
-
- if (bounce_addr == 0)
- bounce_addr =
- (unsigned long)kmap_atomic(sg_page((&bounce_sgl[j])));
-
- while (destlen) {
- src = bounce_addr + bounce_sgl[j].offset;
- srclen = bounce_sgl[j].length - bounce_sgl[j].offset;
-
- copylen = min(srclen, destlen);
- memcpy((void *)dest, (void *)src, copylen);
-
- total_copied += copylen;
- bounce_sgl[j].offset += copylen;
- destlen -= copylen;
- dest += copylen;
-
- if (bounce_sgl[j].offset == bounce_sgl[j].length) {
- /* full */
- kunmap_atomic((void *)bounce_addr);
- j++;
+ vstor_packet->vm_srb.sense_info_length = STORVSC_SENSE_BUFFER_SIZE;
- /*
- * It is possible that the number of elements
- * in the bounce buffer may not be equal to
- * the number of elements in the original
- * scatter list. Handle this correctly.
- */
- if (j == bounce_sgl_count) {
- /*
- * We are done; cleanup and return.
- */
- kunmap_atomic((void *)(dest_addr -
- orig_sgl[i].offset),
- KM_IRQ0);
- local_irq_restore(flags);
- return total_copied;
- }
+ vstor_packet->vm_srb.data_transfer_length =
+ request->data_buffer.len;
- /* if we need to use another bounce buffer */
- if (destlen || i != orig_sgl_count - 1)
- bounce_addr =
- (unsigned long)kmap_atomic(
- sg_page((&bounce_sgl[j])));
- } else if (destlen == 0 && i == orig_sgl_count - 1) {
- /* unmap the last bounce that is < PAGE_SIZE */
- kunmap_atomic((void *)bounce_addr);
- }
- }
+ vstor_packet->operation = VSTOR_OPERATION_EXECUTE_SRB;
- kunmap_atomic((void *)(dest_addr - orig_sgl[i].offset));
+ if (request->data_buffer.len) {
+ ret = vmbus_sendpacket_multipagebuffer(device->channel,
+ &request->data_buffer,
+ vstor_packet,
+ sizeof(struct vstor_packet),
+ (unsigned long)request);
+ } else {
+ ret = vmbus_sendpacket(device->channel, vstor_packet,
+ sizeof(struct vstor_packet),
+ (unsigned long)request,
+ VM_PKT_DATA_INBAND,
+ VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
}
- local_irq_restore(flags);
+ if (ret != 0)
+ return ret;
- return total_copied;
-}
+ atomic_inc(&stor_device->num_outstanding_req);
+ return ret;
+}
-/* Assume the bounce_sgl has enough room ie using the create_bounce_buffer() */
-static unsigned int copy_to_bounce_buffer(struct scatterlist *orig_sgl,
- struct scatterlist *bounce_sgl,
- unsigned int orig_sgl_count)
+static int storvsc_device_alloc(struct scsi_device *sdevice)
{
- int i;
- int j = 0;
- unsigned long src, dest;
- unsigned int srclen, destlen, copylen;
- unsigned int total_copied = 0;
- unsigned long bounce_addr = 0;
- unsigned long src_addr = 0;
- unsigned long flags;
-
- local_irq_save(flags);
+ struct stor_mem_pools *memp;
+ int number = STORVSC_MIN_BUF_NR;
- for (i = 0; i < orig_sgl_count; i++) {
- src_addr = (unsigned long)kmap_atomic(sg_page((&orig_sgl[i])))
- + orig_sgl[i].offset;
- src = src_addr;
- srclen = orig_sgl[i].length;
+ memp = kzalloc(sizeof(struct stor_mem_pools), GFP_KERNEL);
+ if (!memp)
+ return -ENOMEM;
- if (bounce_addr == 0)
- bounce_addr =
- (unsigned long)kmap_atomic(sg_page((&bounce_sgl[j])));
+ memp->request_pool =
+ kmem_cache_create(dev_name(&sdevice->sdev_dev),
+ sizeof(struct storvsc_cmd_request), 0,
+ SLAB_HWCACHE_ALIGN, NULL);
- while (srclen) {
- /* assume bounce offset always == 0 */
- dest = bounce_addr + bounce_sgl[j].length;
- destlen = PAGE_SIZE - bounce_sgl[j].length;
+ if (!memp->request_pool)
+ goto err0;
- copylen = min(srclen, destlen);
- memcpy((void *)dest, (void *)src, copylen);
+ memp->request_mempool = mempool_create(number, mempool_alloc_slab,
+ mempool_free_slab,
+ memp->request_pool);
- total_copied += copylen;
- bounce_sgl[j].length += copylen;
- srclen -= copylen;
- src += copylen;
+ if (!memp->request_mempool)
+ goto err1;
- if (bounce_sgl[j].length == PAGE_SIZE) {
- /* full..move to next entry */
- kunmap_atomic((void *)bounce_addr);
- j++;
+ sdevice->hostdata = memp;
- /* if we need to use another bounce buffer */
- if (srclen || i != orig_sgl_count - 1)
- bounce_addr =
- (unsigned long)kmap_atomic(
- sg_page((&bounce_sgl[j])));
+ return 0;
- } else if (srclen == 0 && i == orig_sgl_count - 1) {
- /* unmap the last bounce that is < PAGE_SIZE */
- kunmap_atomic((void *)bounce_addr);
- }
- }
+err1:
+ kmem_cache_destroy(memp->request_pool);
- kunmap_atomic((void *)(src_addr - orig_sgl[i].offset));
- }
+err0:
+ kfree(memp);
+ return -ENOMEM;
+}
- local_irq_restore(flags);
+static void storvsc_device_destroy(struct scsi_device *sdevice)
+{
+ struct stor_mem_pools *memp = sdevice->hostdata;
- return total_copied;
+ mempool_destroy(memp->request_mempool);
+ kmem_cache_destroy(memp->request_pool);
+ kfree(memp);
+ sdevice->hostdata = NULL;
}
-
-static int storvsc_remove(struct hv_device *dev)
+static int storvsc_device_configure(struct scsi_device *sdevice)
{
- struct storvsc_device *stor_device = hv_get_drvdata(dev);
- struct Scsi_Host *host = stor_device->host;
-
- scsi_remove_host(host);
+ scsi_adjust_queue_depth(sdevice, MSG_SIMPLE_TAG,
+ STORVSC_MAX_IO_REQUESTS);
- scsi_host_put(host);
+ blk_queue_max_segment_size(sdevice->request_queue, PAGE_SIZE);
- storvsc_dev_remove(dev);
+ blk_queue_bounce_limit(sdevice->request_queue, BLK_BOUNCE_ANY);
return 0;
}
-
static int storvsc_get_chs(struct scsi_device *sdev, struct block_device * bdev,
sector_t capacity, int *info)
{
return 0;
}
-static int storvsc_host_reset(struct hv_device *device)
+static int storvsc_host_reset_handler(struct scsi_cmnd *scmnd)
{
+ struct hv_host_device *host_dev = shost_priv(scmnd->device->host);
+ struct hv_device *device = host_dev->dev;
+
struct storvsc_device *stor_device;
- struct hv_storvsc_request *request;
+ struct storvsc_cmd_request *request;
struct vstor_packet *vstor_packet;
int ret, t;
return SUCCESS;
}
-
-/*
- * storvsc_host_reset_handler - Reset the scsi HBA
- */
-static int storvsc_host_reset_handler(struct scsi_cmnd *scmnd)
-{
- struct hv_host_device *host_dev = shost_priv(scmnd->device->host);
- struct hv_device *dev = host_dev->dev;
-
- return storvsc_host_reset(dev);
-}
-
-
-/*
- * storvsc_command_completion - Command completion processing
- */
-static void storvsc_command_completion(struct hv_storvsc_request *request)
-{
- struct storvsc_cmd_request *cmd_request =
- (struct storvsc_cmd_request *)request->context;
- struct scsi_cmnd *scmnd = cmd_request->cmd;
- struct hv_host_device *host_dev = shost_priv(scmnd->device->host);
- void (*scsi_done_fn)(struct scsi_cmnd *);
- struct scsi_sense_hdr sense_hdr;
- struct vmscsi_request *vm_srb;
- struct storvsc_scan_work *wrk;
- struct stor_mem_pools *memp = scmnd->device->hostdata;
-
- vm_srb = &request->vstor_packet.vm_srb;
- if (cmd_request->bounce_sgl_count) {
- if (vm_srb->data_in == READ_TYPE)
- copy_from_bounce_buffer(scsi_sglist(scmnd),
- cmd_request->bounce_sgl,
- scsi_sg_count(scmnd),
- cmd_request->bounce_sgl_count);
- destroy_bounce_buffer(cmd_request->bounce_sgl,
- cmd_request->bounce_sgl_count);
- }
-
- /*
- * If there is an error; offline the device since all
- * error recovery strategies would have already been
- * deployed on the host side.
- */
- if (vm_srb->srb_status == 0x4)
- scmnd->result = DID_TARGET_FAILURE << 16;
- else
- scmnd->result = vm_srb->scsi_status;
-
- /*
- * If the LUN is invalid; remove the device.
- */
- if (vm_srb->srb_status == 0x20) {
- struct storvsc_device *stor_dev;
- struct hv_device *dev = host_dev->dev;
- struct Scsi_Host *host;
-
- stor_dev = get_in_stor_device(dev);
- host = stor_dev->host;
-
- wrk = kmalloc(sizeof(struct storvsc_scan_work),
- GFP_ATOMIC);
- if (!wrk) {
- scmnd->result = DID_TARGET_FAILURE << 16;
- } else {
- wrk->host = host;
- wrk->lun = vm_srb->lun;
- INIT_WORK(&wrk->work, storvsc_remove_lun);
- schedule_work(&wrk->work);
- }
- }
-
- if (scmnd->result) {
- if (scsi_normalize_sense(scmnd->sense_buffer,
- SCSI_SENSE_BUFFERSIZE, &sense_hdr))
- scsi_print_sense_hdr("storvsc", &sense_hdr);
- }
-
- scsi_set_resid(scmnd,
- request->data_buffer.len -
- vm_srb->data_transfer_length);
-
- scsi_done_fn = scmnd->scsi_done;
-
- scmnd->host_scribble = NULL;
- scmnd->scsi_done = NULL;
-
- scsi_done_fn(scmnd);
-
- mempool_free(cmd_request, memp->request_mempool);
-}
-
-static bool storvsc_check_scsi_cmd(struct scsi_cmnd *scmnd)
+static bool storvsc_scsi_cmd_ok(struct scsi_cmnd *scmnd)
{
bool allowed = true;
u8 scsi_op = scmnd->cmnd[0];
switch (scsi_op) {
- /* smartd sends this command, which will offline the device */
+ /*
+ * smartd sends this command and the host does not handle
+ * this. So, don't send it.
+ */
case SET_WINDOW:
scmnd->result = ILLEGAL_REQUEST << 16;
allowed = false;
return allowed;
}
-/*
- * storvsc_queuecommand - Initiate command processing
- */
static int storvsc_queuecommand(struct Scsi_Host *host, struct scsi_cmnd *scmnd)
{
int ret;
struct hv_host_device *host_dev = shost_priv(host);
struct hv_device *dev = host_dev->dev;
- struct hv_storvsc_request *request;
struct storvsc_cmd_request *cmd_request;
unsigned int request_size = 0;
int i;
struct vmscsi_request *vm_srb;
struct stor_mem_pools *memp = scmnd->device->hostdata;
- if (storvsc_check_scsi_cmd(scmnd) == false) {
+ if (!storvsc_scsi_cmd_ok(scmnd)) {
scmnd->scsi_done(scmnd);
return 0;
}
- /* If retrying, no need to prep the cmd */
- if (scmnd->host_scribble) {
-
- cmd_request =
- (struct storvsc_cmd_request *)scmnd->host_scribble;
-
- goto retry_request;
- }
-
request_size = sizeof(struct storvsc_cmd_request);
cmd_request = mempool_alloc(memp->request_mempool,
GFP_ATOMIC);
+
+ /*
+ * We might be invoked in an interrupt context; hence
+ * mempool_alloc() can fail.
+ */
if (!cmd_request)
return SCSI_MLQUEUE_DEVICE_BUSY;
memset(cmd_request, 0, sizeof(struct storvsc_cmd_request));
/* Setup the cmd request */
- cmd_request->bounce_sgl_count = 0;
- cmd_request->bounce_sgl = NULL;
cmd_request->cmd = scmnd;
scmnd->host_scribble = (unsigned char *)cmd_request;
- request = &cmd_request->request;
- vm_srb = &request->vstor_packet.vm_srb;
+ vm_srb = &cmd_request->vstor_packet.vm_srb;
/* Build the SRB */
break;
}
- request->on_io_completion = storvsc_command_completion;
- request->context = cmd_request;/* scmnd; */
vm_srb->port_number = host_dev->port;
vm_srb->path_id = scmnd->device->channel;
memcpy(vm_srb->cdb, scmnd->cmnd, vm_srb->cdb_length);
- request->sense_buffer = scmnd->sense_buffer;
+ cmd_request->sense_buffer = scmnd->sense_buffer;
- request->data_buffer.len = scsi_bufflen(scmnd);
+ cmd_request->data_buffer.len = scsi_bufflen(scmnd);
if (scsi_sg_count(scmnd)) {
sgl = (struct scatterlist *)scsi_sglist(scmnd);
sg_count = scsi_sg_count(scmnd);
scsi_bufflen(scmnd),
vm_srb->data_in);
if (!cmd_request->bounce_sgl) {
- scmnd->host_scribble = NULL;
- mempool_free(cmd_request,
- memp->request_mempool);
-
- return SCSI_MLQUEUE_HOST_BUSY;
+ ret = SCSI_MLQUEUE_HOST_BUSY;
+ goto queue_error;
}
cmd_request->bounce_sgl_count =
sg_count = cmd_request->bounce_sgl_count;
}
- request->data_buffer.offset = sgl[0].offset;
+ cmd_request->data_buffer.offset = sgl[0].offset;
for (i = 0; i < sg_count; i++)
- request->data_buffer.pfn_array[i] =
+ cmd_request->data_buffer.pfn_array[i] =
page_to_pfn(sg_page((&sgl[i])));
} else if (scsi_sglist(scmnd)) {
- request->data_buffer.offset =
+ cmd_request->data_buffer.offset =
virt_to_phys(scsi_sglist(scmnd)) & (PAGE_SIZE-1);
- request->data_buffer.pfn_array[0] =
+ cmd_request->data_buffer.pfn_array[0] =
virt_to_phys(scsi_sglist(scmnd)) >> PAGE_SHIFT;
}
-retry_request:
/* Invokes the vsc to start an IO */
- ret = storvsc_do_io(dev, &cmd_request->request);
+ ret = storvsc_do_io(dev, cmd_request);
if (ret == -EAGAIN) {
/* no more space */
- if (cmd_request->bounce_sgl_count)
+ if (cmd_request->bounce_sgl_count) {
destroy_bounce_buffer(cmd_request->bounce_sgl,
cmd_request->bounce_sgl_count);
- mempool_free(cmd_request, memp->request_mempool);
-
- scmnd->host_scribble = NULL;
-
- ret = SCSI_MLQUEUE_DEVICE_BUSY;
+ ret = SCSI_MLQUEUE_DEVICE_BUSY;
+ goto queue_error;
+ }
}
+ return 0;
+
+queue_error:
+ mempool_free(cmd_request, memp->request_mempool);
+ scmnd->host_scribble = NULL;
return ret;
}
-/* Scsi driver */
static struct scsi_host_template scsi_driver = {
.module = THIS_MODULE,
.name = "storvsc_host_t",
MODULE_DEVICE_TABLE(vmbus, id_table);
-
-/*
- * storvsc_probe - Add a new device for this driver
- */
-
static int storvsc_probe(struct hv_device *device,
const struct hv_vmbus_device_id *dev_id)
{
struct Scsi_Host *host;
struct hv_host_device *host_dev;
bool dev_is_ide = ((dev_id->driver_data == IDE_GUID) ? true : false);
- int path = 0;
int target = 0;
struct storvsc_device *stor_device;
if (ret)
goto err_out1;
- if (dev_is_ide)
- storvsc_get_ide_info(device, &target, &path);
-
host_dev->path = stor_device->path_id;
host_dev->target = stor_device->target_id;
if (!dev_is_ide) {
scsi_scan_host(host);
- return 0;
- }
- ret = scsi_add_device(host, 0, target, 0);
- if (ret) {
- scsi_remove_host(host);
- goto err_out2;
+ } else {
+ target = (device->dev_instance.b[5] << 8 |
+ device->dev_instance.b[4]);
+ ret = scsi_add_device(host, 0, target, 0);
+ if (ret) {
+ scsi_remove_host(host);
+ goto err_out2;
+ }
}
return 0;
return ret;
}
-/* The one and only one */
+static int storvsc_remove(struct hv_device *dev)
+{
+ struct storvsc_device *stor_device = hv_get_drvdata(dev);
+ struct Scsi_Host *host = stor_device->host;
+
+ scsi_remove_host(host);
+ storvsc_dev_remove(dev);
+ scsi_host_put(host);
+
+ return 0;
+}
static struct hv_driver storvsc_drv = {
.name = KBUILD_MODNAME,
* This is called from xv_malloc/xv_free path, so it
* needs to be fast.
*/
- static void *get_ptr_atomic(struct page *page, u16 offset, enum km_type type)
+ static void *get_ptr_atomic(struct page *page, u16 offset)
{
unsigned char *base;
- base = kmap_atomic(page, type);
+ base = kmap_atomic(page);
return base + offset;
}
- static void put_ptr_atomic(void *ptr, enum km_type type)
+ static void put_ptr_atomic(void *ptr)
{
- kunmap_atomic(ptr, type);
+ kunmap_atomic(ptr);
}
static u32 get_blockprev(struct block_header *block)
if (block->link.next_page) {
nextblock = get_ptr_atomic(block->link.next_page,
- block->link.next_offset, KM_USER1);
+ block->link.next_offset);
nextblock->link.prev_page = page;
nextblock->link.prev_offset = offset;
- put_ptr_atomic(nextblock, KM_USER1);
+ put_ptr_atomic(nextblock);
/* If there was a next page then the free bits are set. */
return;
}
if (block->link.prev_page) {
tmpblock = get_ptr_atomic(block->link.prev_page,
- block->link.prev_offset, KM_USER1);
+ block->link.prev_offset);
tmpblock->link.next_page = block->link.next_page;
tmpblock->link.next_offset = block->link.next_offset;
- put_ptr_atomic(tmpblock, KM_USER1);
+ put_ptr_atomic(tmpblock);
}
if (block->link.next_page) {
tmpblock = get_ptr_atomic(block->link.next_page,
- block->link.next_offset, KM_USER1);
+ block->link.next_offset);
tmpblock->link.prev_page = block->link.prev_page;
tmpblock->link.prev_offset = block->link.prev_offset;
- put_ptr_atomic(tmpblock, KM_USER1);
+ put_ptr_atomic(tmpblock);
}
/* Is this block is at the head of the freelist? */
if (pool->freelist[slindex].page) {
struct block_header *tmpblock;
tmpblock = get_ptr_atomic(pool->freelist[slindex].page,
- pool->freelist[slindex].offset,
- KM_USER1);
+ pool->freelist[slindex].offset);
tmpblock->link.prev_page = NULL;
tmpblock->link.prev_offset = 0;
- put_ptr_atomic(tmpblock, KM_USER1);
+ put_ptr_atomic(tmpblock);
} else {
/* This freelist bucket is empty */
__clear_bit(slindex % BITS_PER_LONG,
stat_inc(&pool->total_pages);
spin_lock(&pool->lock);
- block = get_ptr_atomic(page, 0, KM_USER0);
+ block = get_ptr_atomic(page, 0);
block->size = PAGE_SIZE - XV_ALIGN;
set_flag(block, BLOCK_FREE);
insert_block(pool, page, 0, block);
- put_ptr_atomic(block, KM_USER0);
+ put_ptr_atomic(block);
spin_unlock(&pool->lock);
return 0;
return -ENOMEM;
}
- block = get_ptr_atomic(*page, *offset, KM_USER0);
+ block = get_ptr_atomic(*page, *offset);
remove_block(pool, *page, *offset, block, index);
block->size = origsize;
clear_flag(block, BLOCK_FREE);
- put_ptr_atomic(block, KM_USER0);
+ put_ptr_atomic(block);
spin_unlock(&pool->lock);
*offset += XV_ALIGN;
spin_lock(&pool->lock);
- page_start = get_ptr_atomic(page, 0, KM_USER0);
+ page_start = get_ptr_atomic(page, 0);
block = (struct block_header *)((char *)page_start + offset);
/* Catch double free bugs */
/* No used objects in this page. Free it. */
if (block->size == PAGE_SIZE - XV_ALIGN) {
- put_ptr_atomic(page_start, KM_USER0);
+ put_ptr_atomic(page_start);
spin_unlock(&pool->lock);
__free_page(page);
set_blockprev(tmpblock, offset);
}
- put_ptr_atomic(page_start, KM_USER0);
+ put_ptr_atomic(page_start);
spin_unlock(&pool->lock);
}
EXPORT_SYMBOL_GPL(xv_free);
--- /dev/null
- to_va = kmap_atomic(page, KM_USER0);
+/*
+ * zcache.c
+ *
+ * Copyright (c) 2010-2012, Dan Magenheimer, Oracle Corp.
+ * Copyright (c) 2010,2011, Nitin Gupta
+ *
+ * Zcache provides an in-kernel "host implementation" for transcendent memory
+ * and, thus indirectly, for cleancache and frontswap. Zcache includes two
+ * page-accessible memory [1] interfaces, both utilizing lzo1x compression:
+ * 1) "compression buddies" ("zbud") is used for ephemeral pages
+ * 2) xvmalloc is used for persistent pages.
+ * Xvmalloc (based on the TLSF allocator) has very low fragmentation
+ * so maximizes space efficiency, while zbud allows pairs (and potentially,
+ * in the future, more than a pair of) compressed pages to be closely linked
+ * so that reclaiming can be done via the kernel's physical-page-oriented
+ * "shrinker" interface.
+ *
+ * [1] For a definition of page-accessible memory (aka PAM), see:
+ * http://marc.info/?l=linux-mm&m=127811271605009
+ * RAMSTER TODO:
+ * - handle remotifying of buddied pages (see zbud_remotify_zbpg)
+ * - kernel boot params: nocleancache/nofrontswap don't always work?!?
+ */
+
+#include <linux/module.h>
+#include <linux/cpu.h>
+#include <linux/highmem.h>
+#include <linux/list.h>
+#include <linux/lzo.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/types.h>
+#include <linux/atomic.h>
+#include <linux/math64.h>
+#include "tmem.h"
+#include "zcache.h"
+#include "ramster.h"
+#include "cluster/tcp.h"
+
+#include "xvmalloc.h" /* temporary until change to zsmalloc */
+
+#define RAMSTER_TESTING
+
+#if (!defined(CONFIG_CLEANCACHE) && !defined(CONFIG_FRONTSWAP))
+#error "ramster is useless without CONFIG_CLEANCACHE or CONFIG_FRONTSWAP"
+#endif
+#ifdef CONFIG_CLEANCACHE
+#include <linux/cleancache.h>
+#endif
+#ifdef CONFIG_FRONTSWAP
+#include <linux/frontswap.h>
+#endif
+
+enum ramster_remotify_op {
+ RAMSTER_REMOTIFY_EPH_PUT,
+ RAMSTER_REMOTIFY_PERS_PUT,
+ RAMSTER_REMOTIFY_FLUSH_PAGE,
+ RAMSTER_REMOTIFY_FLUSH_OBJ,
+ RAMSTER_INTRANSIT_PERS
+};
+
+struct ramster_remotify_hdr {
+ enum ramster_remotify_op op;
+ struct list_head list;
+};
+
+#define ZBH_SENTINEL 0x43214321
+#define ZBPG_SENTINEL 0xdeadbeef
+
+#define ZBUD_MAX_BUDS 2
+
+struct zbud_hdr {
+ struct ramster_remotify_hdr rem_op;
+ uint16_t client_id;
+ uint16_t pool_id;
+ struct tmem_oid oid;
+ uint32_t index;
+ uint16_t size; /* compressed size in bytes, zero means unused */
+ DECL_SENTINEL
+};
+
+#define ZVH_SENTINEL 0x43214321
+static const int zv_max_page_size = (PAGE_SIZE / 8) * 7;
+
+struct zv_hdr {
+ struct ramster_remotify_hdr rem_op;
+ uint16_t client_id;
+ uint16_t pool_id;
+ struct tmem_oid oid;
+ uint32_t index;
+ DECL_SENTINEL
+};
+
+struct flushlist_node {
+ struct ramster_remotify_hdr rem_op;
+ struct tmem_xhandle xh;
+};
+
+union {
+ struct ramster_remotify_hdr rem_op;
+ struct zv_hdr zv;
+ struct zbud_hdr zbud;
+ struct flushlist_node flist;
+} remotify_list_node;
+
+static LIST_HEAD(zcache_rem_op_list);
+static DEFINE_SPINLOCK(zcache_rem_op_list_lock);
+
+#if 0
+/* this is more aggressive but may cause other problems? */
+#define ZCACHE_GFP_MASK (GFP_ATOMIC | __GFP_NORETRY | __GFP_NOWARN)
+#else
+#define ZCACHE_GFP_MASK \
+ (__GFP_FS | __GFP_NORETRY | __GFP_NOWARN | __GFP_NOMEMALLOC)
+#endif
+
+#define MAX_POOLS_PER_CLIENT 16
+
+#define MAX_CLIENTS 16
+#define LOCAL_CLIENT ((uint16_t)-1)
+
+MODULE_LICENSE("GPL");
+
+struct zcache_client {
+ struct tmem_pool *tmem_pools[MAX_POOLS_PER_CLIENT];
+ struct xv_pool *xvpool;
+ bool allocated;
+ atomic_t refcount;
+};
+
+static struct zcache_client zcache_host;
+static struct zcache_client zcache_clients[MAX_CLIENTS];
+
+static inline uint16_t get_client_id_from_client(struct zcache_client *cli)
+{
+ BUG_ON(cli == NULL);
+ if (cli == &zcache_host)
+ return LOCAL_CLIENT;
+ return cli - &zcache_clients[0];
+}
+
+static inline bool is_local_client(struct zcache_client *cli)
+{
+ return cli == &zcache_host;
+}
+
+/**********
+ * Compression buddies ("zbud") provides for packing two (or, possibly
+ * in the future, more) compressed ephemeral pages into a single "raw"
+ * (physical) page and tracking them with data structures so that
+ * the raw pages can be easily reclaimed.
+ *
+ * A zbud page ("zbpg") is an aligned page containing a list_head,
+ * a lock, and two "zbud headers". The remainder of the physical
+ * page is divided up into aligned 64-byte "chunks" which contain
+ * the compressed data for zero, one, or two zbuds. Each zbpg
+ * resides on: (1) an "unused list" if it has no zbuds; (2) a
+ * "buddied" list if it is fully populated with two zbuds; or
+ * (3) one of PAGE_SIZE/64 "unbuddied" lists indexed by how many chunks
+ * the one unbuddied zbud uses. The data inside a zbpg cannot be
+ * read or written unless the zbpg's lock is held.
+ */
+
+struct zbud_page {
+ struct list_head bud_list;
+ spinlock_t lock;
+ struct zbud_hdr buddy[ZBUD_MAX_BUDS];
+ DECL_SENTINEL
+ /* followed by NUM_CHUNK aligned CHUNK_SIZE-byte chunks */
+};
+
+#define CHUNK_SHIFT 6
+#define CHUNK_SIZE (1 << CHUNK_SHIFT)
+#define CHUNK_MASK (~(CHUNK_SIZE-1))
+#define NCHUNKS (((PAGE_SIZE - sizeof(struct zbud_page)) & \
+ CHUNK_MASK) >> CHUNK_SHIFT)
+#define MAX_CHUNK (NCHUNKS-1)
+
+static struct {
+ struct list_head list;
+ unsigned count;
+} zbud_unbuddied[NCHUNKS];
+/* list N contains pages with N chunks USED and NCHUNKS-N unused */
+/* element 0 is never used but optimizing that isn't worth it */
+static unsigned long zbud_cumul_chunk_counts[NCHUNKS];
+
+struct list_head zbud_buddied_list;
+static unsigned long zcache_zbud_buddied_count;
+
+/* protects the buddied list and all unbuddied lists */
+static DEFINE_SPINLOCK(zbud_budlists_spinlock);
+
+static atomic_t zcache_zbud_curr_raw_pages;
+static atomic_t zcache_zbud_curr_zpages;
+static unsigned long zcache_zbud_curr_zbytes;
+static unsigned long zcache_zbud_cumul_zpages;
+static unsigned long zcache_zbud_cumul_zbytes;
+static unsigned long zcache_compress_poor;
+static unsigned long zcache_policy_percent_exceeded;
+static unsigned long zcache_mean_compress_poor;
+
+/*
+ * RAMster counters
+ * - Remote pages are pages with a local pampd but the data is remote
+ * - Foreign pages are pages stored locally but belonging to another node
+ */
+static atomic_t ramster_remote_pers_pages = ATOMIC_INIT(0);
+static unsigned long ramster_pers_remotify_enable;
+static unsigned long ramster_eph_remotify_enable;
+static unsigned long ramster_eph_pages_remoted;
+static unsigned long ramster_eph_pages_remote_failed;
+static unsigned long ramster_pers_pages_remoted;
+static unsigned long ramster_pers_pages_remote_failed;
+static unsigned long ramster_pers_pages_remote_nomem;
+static unsigned long ramster_remote_objects_flushed;
+static unsigned long ramster_remote_object_flushes_failed;
+static unsigned long ramster_remote_pages_flushed;
+static unsigned long ramster_remote_page_flushes_failed;
+static unsigned long ramster_remote_eph_pages_succ_get;
+static unsigned long ramster_remote_pers_pages_succ_get;
+static unsigned long ramster_remote_eph_pages_unsucc_get;
+static unsigned long ramster_remote_pers_pages_unsucc_get;
+static atomic_t ramster_curr_flnode_count = ATOMIC_INIT(0);
+static unsigned long ramster_curr_flnode_count_max;
+static atomic_t ramster_foreign_eph_pampd_count = ATOMIC_INIT(0);
+static unsigned long ramster_foreign_eph_pampd_count_max;
+static atomic_t ramster_foreign_pers_pampd_count = ATOMIC_INIT(0);
+static unsigned long ramster_foreign_pers_pampd_count_max;
+
+/* forward references */
+static void *zcache_get_free_page(void);
+static void zcache_free_page(void *p);
+
+/*
+ * zbud helper functions
+ */
+
+static inline unsigned zbud_max_buddy_size(void)
+{
+ return MAX_CHUNK << CHUNK_SHIFT;
+}
+
+static inline unsigned zbud_size_to_chunks(unsigned size)
+{
+ BUG_ON(size == 0 || size > zbud_max_buddy_size());
+ return (size + CHUNK_SIZE - 1) >> CHUNK_SHIFT;
+}
+
+static inline int zbud_budnum(struct zbud_hdr *zh)
+{
+ unsigned offset = (unsigned long)zh & (PAGE_SIZE - 1);
+ struct zbud_page *zbpg = NULL;
+ unsigned budnum = -1U;
+ int i;
+
+ for (i = 0; i < ZBUD_MAX_BUDS; i++)
+ if (offset == offsetof(typeof(*zbpg), buddy[i])) {
+ budnum = i;
+ break;
+ }
+ BUG_ON(budnum == -1U);
+ return budnum;
+}
+
+static char *zbud_data(struct zbud_hdr *zh, unsigned size)
+{
+ struct zbud_page *zbpg;
+ char *p;
+ unsigned budnum;
+
+ ASSERT_SENTINEL(zh, ZBH);
+ budnum = zbud_budnum(zh);
+ BUG_ON(size == 0 || size > zbud_max_buddy_size());
+ zbpg = container_of(zh, struct zbud_page, buddy[budnum]);
+ ASSERT_SPINLOCK(&zbpg->lock);
+ p = (char *)zbpg;
+ if (budnum == 0)
+ p += ((sizeof(struct zbud_page) + CHUNK_SIZE - 1) &
+ CHUNK_MASK);
+ else if (budnum == 1)
+ p += PAGE_SIZE - ((size + CHUNK_SIZE - 1) & CHUNK_MASK);
+ return p;
+}
+
+static void zbud_copy_from_pampd(char *data, size_t *size, struct zbud_hdr *zh)
+{
+ struct zbud_page *zbpg;
+ char *p;
+ unsigned budnum;
+
+ ASSERT_SENTINEL(zh, ZBH);
+ budnum = zbud_budnum(zh);
+ zbpg = container_of(zh, struct zbud_page, buddy[budnum]);
+ spin_lock(&zbpg->lock);
+ BUG_ON(zh->size > *size);
+ p = (char *)zbpg;
+ if (budnum == 0)
+ p += ((sizeof(struct zbud_page) + CHUNK_SIZE - 1) &
+ CHUNK_MASK);
+ else if (budnum == 1)
+ p += PAGE_SIZE - ((zh->size + CHUNK_SIZE - 1) & CHUNK_MASK);
+ /* client should be filled in by caller */
+ memcpy(data, p, zh->size);
+ *size = zh->size;
+ spin_unlock(&zbpg->lock);
+}
+
+/*
+ * zbud raw page management
+ */
+
+static struct zbud_page *zbud_alloc_raw_page(void)
+{
+ struct zbud_page *zbpg = NULL;
+ struct zbud_hdr *zh0, *zh1;
+ zbpg = zcache_get_free_page();
+ if (likely(zbpg != NULL)) {
+ INIT_LIST_HEAD(&zbpg->bud_list);
+ zh0 = &zbpg->buddy[0]; zh1 = &zbpg->buddy[1];
+ spin_lock_init(&zbpg->lock);
+ atomic_inc(&zcache_zbud_curr_raw_pages);
+ INIT_LIST_HEAD(&zbpg->bud_list);
+ SET_SENTINEL(zbpg, ZBPG);
+ zh0->size = 0; zh1->size = 0;
+ tmem_oid_set_invalid(&zh0->oid);
+ tmem_oid_set_invalid(&zh1->oid);
+ }
+ return zbpg;
+}
+
+static void zbud_free_raw_page(struct zbud_page *zbpg)
+{
+ struct zbud_hdr *zh0 = &zbpg->buddy[0], *zh1 = &zbpg->buddy[1];
+
+ ASSERT_SENTINEL(zbpg, ZBPG);
+ BUG_ON(!list_empty(&zbpg->bud_list));
+ ASSERT_SPINLOCK(&zbpg->lock);
+ BUG_ON(zh0->size != 0 || tmem_oid_valid(&zh0->oid));
+ BUG_ON(zh1->size != 0 || tmem_oid_valid(&zh1->oid));
+ INVERT_SENTINEL(zbpg, ZBPG);
+ spin_unlock(&zbpg->lock);
+ atomic_dec(&zcache_zbud_curr_raw_pages);
+ zcache_free_page(zbpg);
+}
+
+/*
+ * core zbud handling routines
+ */
+
+static unsigned zbud_free(struct zbud_hdr *zh)
+{
+ unsigned size;
+
+ ASSERT_SENTINEL(zh, ZBH);
+ BUG_ON(!tmem_oid_valid(&zh->oid));
+ size = zh->size;
+ BUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());
+ zh->size = 0;
+ tmem_oid_set_invalid(&zh->oid);
+ INVERT_SENTINEL(zh, ZBH);
+ zcache_zbud_curr_zbytes -= size;
+ atomic_dec(&zcache_zbud_curr_zpages);
+ return size;
+}
+
+static void zbud_free_and_delist(struct zbud_hdr *zh)
+{
+ unsigned chunks;
+ struct zbud_hdr *zh_other;
+ unsigned budnum = zbud_budnum(zh), size;
+ struct zbud_page *zbpg =
+ container_of(zh, struct zbud_page, buddy[budnum]);
+
+ /* FIXME, should be BUG_ON, pool destruction path doesn't disable
+ * interrupts tmem_destroy_pool()->tmem_pampd_destroy_all_in_obj()->
+ * tmem_objnode_node_destroy()-> zcache_pampd_free() */
+ WARN_ON(!irqs_disabled());
+ spin_lock(&zbpg->lock);
+ if (list_empty(&zbpg->bud_list)) {
+ /* ignore zombie page... see zbud_evict_pages() */
+ spin_unlock(&zbpg->lock);
+ return;
+ }
+ size = zbud_free(zh);
+ ASSERT_SPINLOCK(&zbpg->lock);
+ zh_other = &zbpg->buddy[(budnum == 0) ? 1 : 0];
+ if (zh_other->size == 0) { /* was unbuddied: unlist and free */
+ chunks = zbud_size_to_chunks(size) ;
+ spin_lock(&zbud_budlists_spinlock);
+ BUG_ON(list_empty(&zbud_unbuddied[chunks].list));
+ list_del_init(&zbpg->bud_list);
+ zbud_unbuddied[chunks].count--;
+ spin_unlock(&zbud_budlists_spinlock);
+ zbud_free_raw_page(zbpg);
+ } else { /* was buddied: move remaining buddy to unbuddied list */
+ chunks = zbud_size_to_chunks(zh_other->size) ;
+ spin_lock(&zbud_budlists_spinlock);
+ list_del_init(&zbpg->bud_list);
+ zcache_zbud_buddied_count--;
+ list_add_tail(&zbpg->bud_list, &zbud_unbuddied[chunks].list);
+ zbud_unbuddied[chunks].count++;
+ spin_unlock(&zbud_budlists_spinlock);
+ spin_unlock(&zbpg->lock);
+ }
+}
+
+static struct zbud_hdr *zbud_create(uint16_t client_id, uint16_t pool_id,
+ struct tmem_oid *oid,
+ uint32_t index, struct page *page,
+ void *cdata, unsigned size)
+{
+ struct zbud_hdr *zh0, *zh1, *zh = NULL;
+ struct zbud_page *zbpg = NULL, *ztmp;
+ unsigned nchunks;
+ char *to;
+ int i, found_good_buddy = 0;
+
+ nchunks = zbud_size_to_chunks(size) ;
+ for (i = MAX_CHUNK - nchunks + 1; i > 0; i--) {
+ spin_lock(&zbud_budlists_spinlock);
+ if (!list_empty(&zbud_unbuddied[i].list)) {
+ list_for_each_entry_safe(zbpg, ztmp,
+ &zbud_unbuddied[i].list, bud_list) {
+ if (spin_trylock(&zbpg->lock)) {
+ found_good_buddy = i;
+ goto found_unbuddied;
+ }
+ }
+ }
+ spin_unlock(&zbud_budlists_spinlock);
+ }
+ /* didn't find a good buddy, try allocating a new page */
+ zbpg = zbud_alloc_raw_page();
+ if (unlikely(zbpg == NULL))
+ goto out;
+ /* ok, have a page, now compress the data before taking locks */
+ spin_lock(&zbud_budlists_spinlock);
+ spin_lock(&zbpg->lock);
+ list_add_tail(&zbpg->bud_list, &zbud_unbuddied[nchunks].list);
+ zbud_unbuddied[nchunks].count++;
+ zh = &zbpg->buddy[0];
+ goto init_zh;
+
+found_unbuddied:
+ ASSERT_SPINLOCK(&zbpg->lock);
+ zh0 = &zbpg->buddy[0]; zh1 = &zbpg->buddy[1];
+ BUG_ON(!((zh0->size == 0) ^ (zh1->size == 0)));
+ if (zh0->size != 0) { /* buddy0 in use, buddy1 is vacant */
+ ASSERT_SENTINEL(zh0, ZBH);
+ zh = zh1;
+ } else if (zh1->size != 0) { /* buddy1 in use, buddy0 is vacant */
+ ASSERT_SENTINEL(zh1, ZBH);
+ zh = zh0;
+ } else
+ BUG();
+ list_del_init(&zbpg->bud_list);
+ zbud_unbuddied[found_good_buddy].count--;
+ list_add_tail(&zbpg->bud_list, &zbud_buddied_list);
+ zcache_zbud_buddied_count++;
+
+init_zh:
+ SET_SENTINEL(zh, ZBH);
+ zh->size = size;
+ zh->index = index;
+ zh->oid = *oid;
+ zh->pool_id = pool_id;
+ zh->client_id = client_id;
+ to = zbud_data(zh, size);
+ memcpy(to, cdata, size);
+ spin_unlock(&zbpg->lock);
+ spin_unlock(&zbud_budlists_spinlock);
+ zbud_cumul_chunk_counts[nchunks]++;
+ atomic_inc(&zcache_zbud_curr_zpages);
+ zcache_zbud_cumul_zpages++;
+ zcache_zbud_curr_zbytes += size;
+ zcache_zbud_cumul_zbytes += size;
+out:
+ return zh;
+}
+
+static int zbud_decompress(struct page *page, struct zbud_hdr *zh)
+{
+ struct zbud_page *zbpg;
+ unsigned budnum = zbud_budnum(zh);
+ size_t out_len = PAGE_SIZE;
+ char *to_va, *from_va;
+ unsigned size;
+ int ret = 0;
+
+ zbpg = container_of(zh, struct zbud_page, buddy[budnum]);
+ spin_lock(&zbpg->lock);
+ if (list_empty(&zbpg->bud_list)) {
+ /* ignore zombie page... see zbud_evict_pages() */
+ ret = -EINVAL;
+ goto out;
+ }
+ ASSERT_SENTINEL(zh, ZBH);
+ BUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());
- kunmap_atomic(to_va, KM_USER0);
++ to_va = kmap_atomic(page);
+ size = zh->size;
+ from_va = zbud_data(zh, size);
+ ret = lzo1x_decompress_safe(from_va, size, to_va, &out_len);
+ BUG_ON(ret != LZO_E_OK);
+ BUG_ON(out_len != PAGE_SIZE);
- zv = kmap_atomic(page, KM_USER0) + offset;
++ kunmap_atomic(to_va);
+out:
+ spin_unlock(&zbpg->lock);
+ return ret;
+}
+
+/*
+ * The following routines handle shrinking of ephemeral pages by evicting
+ * pages "least valuable" first.
+ */
+
+static unsigned long zcache_evicted_raw_pages;
+static unsigned long zcache_evicted_buddied_pages;
+static unsigned long zcache_evicted_unbuddied_pages;
+
+static struct tmem_pool *zcache_get_pool_by_id(uint16_t cli_id,
+ uint16_t poolid);
+static void zcache_put_pool(struct tmem_pool *pool);
+
+/*
+ * Flush and free all zbuds in a zbpg, then free the pageframe
+ */
+static void zbud_evict_zbpg(struct zbud_page *zbpg)
+{
+ struct zbud_hdr *zh;
+ int i, j;
+ uint32_t pool_id[ZBUD_MAX_BUDS], client_id[ZBUD_MAX_BUDS];
+ uint32_t index[ZBUD_MAX_BUDS];
+ struct tmem_oid oid[ZBUD_MAX_BUDS];
+ struct tmem_pool *pool;
+ unsigned long flags;
+
+ ASSERT_SPINLOCK(&zbpg->lock);
+ for (i = 0, j = 0; i < ZBUD_MAX_BUDS; i++) {
+ zh = &zbpg->buddy[i];
+ if (zh->size) {
+ client_id[j] = zh->client_id;
+ pool_id[j] = zh->pool_id;
+ oid[j] = zh->oid;
+ index[j] = zh->index;
+ j++;
+ }
+ }
+ spin_unlock(&zbpg->lock);
+ for (i = 0; i < j; i++) {
+ pool = zcache_get_pool_by_id(client_id[i], pool_id[i]);
+ BUG_ON(pool == NULL);
+ local_irq_save(flags);
+ /* these flushes should dispose of any local storage */
+ tmem_flush_page(pool, &oid[i], index[i]);
+ local_irq_restore(flags);
+ zcache_put_pool(pool);
+ }
+}
+
+/*
+ * Free nr pages. This code is funky because we want to hold the locks
+ * protecting various lists for as short a time as possible, and in some
+ * circumstances the list may change asynchronously when the list lock is
+ * not held. In some cases we also trylock not only to avoid waiting on a
+ * page in use by another cpu, but also to avoid potential deadlock due to
+ * lock inversion.
+ */
+static void zbud_evict_pages(int nr)
+{
+ struct zbud_page *zbpg;
+ int i, newly_unused_pages = 0;
+
+
+ /* now try freeing unbuddied pages, starting with least space avail */
+ for (i = 0; i < MAX_CHUNK; i++) {
+retry_unbud_list_i:
+ spin_lock_bh(&zbud_budlists_spinlock);
+ if (list_empty(&zbud_unbuddied[i].list)) {
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ continue;
+ }
+ list_for_each_entry(zbpg, &zbud_unbuddied[i].list, bud_list) {
+ if (unlikely(!spin_trylock(&zbpg->lock)))
+ continue;
+ zbud_unbuddied[i].count--;
+ spin_unlock(&zbud_budlists_spinlock);
+ zcache_evicted_unbuddied_pages++;
+ /* want budlists unlocked when doing zbpg eviction */
+ zbud_evict_zbpg(zbpg);
+ newly_unused_pages++;
+ local_bh_enable();
+ if (--nr <= 0)
+ goto evict_unused;
+ goto retry_unbud_list_i;
+ }
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ }
+
+ /* as a last resort, free buddied pages */
+retry_bud_list:
+ spin_lock_bh(&zbud_budlists_spinlock);
+ if (list_empty(&zbud_buddied_list)) {
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ goto evict_unused;
+ }
+ list_for_each_entry(zbpg, &zbud_buddied_list, bud_list) {
+ if (unlikely(!spin_trylock(&zbpg->lock)))
+ continue;
+ zcache_zbud_buddied_count--;
+ spin_unlock(&zbud_budlists_spinlock);
+ zcache_evicted_buddied_pages++;
+ /* want budlists unlocked when doing zbpg eviction */
+ zbud_evict_zbpg(zbpg);
+ newly_unused_pages++;
+ local_bh_enable();
+ if (--nr <= 0)
+ goto evict_unused;
+ goto retry_bud_list;
+ }
+ spin_unlock_bh(&zbud_budlists_spinlock);
+
+evict_unused:
+ return;
+}
+
+static DEFINE_PER_CPU(unsigned char *, zcache_remoteputmem);
+
+static int zbud_remotify_zbud(struct tmem_xhandle *xh, char *data,
+ size_t size)
+{
+ struct tmem_pool *pool;
+ int i, remotenode, ret = -1;
+ unsigned char cksum, *p;
+ unsigned long flags;
+
+ for (p = data, cksum = 0, i = 0; i < size; i++)
+ cksum += *p;
+ ret = ramster_remote_put(xh, data, size, true, &remotenode);
+ if (ret == 0) {
+ /* data was successfully remoted so change the local version
+ * to point to the remote node where it landed */
+ pool = zcache_get_pool_by_id(LOCAL_CLIENT, xh->pool_id);
+ BUG_ON(pool == NULL);
+ local_irq_save(flags);
+ /* tmem_replace will also free up any local space */
+ (void)tmem_replace(pool, &xh->oid, xh->index,
+ pampd_make_remote(remotenode, size, cksum));
+ local_irq_restore(flags);
+ zcache_put_pool(pool);
+ ramster_eph_pages_remoted++;
+ ret = 0;
+ } else
+ ramster_eph_pages_remote_failed++;
+ return ret;
+}
+
+static int zbud_remotify_zbpg(struct zbud_page *zbpg)
+{
+ struct zbud_hdr *zh1, *zh2 = NULL;
+ struct tmem_xhandle xh1, xh2 = { 0 };
+ char *data1 = NULL, *data2 = NULL;
+ size_t size1 = 0, size2 = 0;
+ int ret = 0;
+ unsigned char *tmpmem = __get_cpu_var(zcache_remoteputmem);
+
+ ASSERT_SPINLOCK(&zbpg->lock);
+ if (zbpg->buddy[0].size == 0)
+ zh1 = &zbpg->buddy[1];
+ else if (zbpg->buddy[1].size == 0)
+ zh1 = &zbpg->buddy[0];
+ else {
+ zh1 = &zbpg->buddy[0];
+ zh2 = &zbpg->buddy[1];
+ }
+ /* don't remotify pages that are already remotified */
+ if (zh1->client_id != LOCAL_CLIENT)
+ zh1 = NULL;
+ if ((zh2 != NULL) && (zh2->client_id != LOCAL_CLIENT))
+ zh2 = NULL;
+
+ /* copy the data and metadata so can release lock */
+ if (zh1 != NULL) {
+ xh1.client_id = zh1->client_id;
+ xh1.pool_id = zh1->pool_id;
+ xh1.oid = zh1->oid;
+ xh1.index = zh1->index;
+ size1 = zh1->size;
+ data1 = zbud_data(zh1, size1);
+ memcpy(tmpmem, zbud_data(zh1, size1), size1);
+ data1 = tmpmem;
+ tmpmem += size1;
+ }
+ if (zh2 != NULL) {
+ xh2.client_id = zh2->client_id;
+ xh2.pool_id = zh2->pool_id;
+ xh2.oid = zh2->oid;
+ xh2.index = zh2->index;
+ size2 = zh2->size;
+ memcpy(tmpmem, zbud_data(zh2, size2), size2);
+ data2 = tmpmem;
+ }
+ spin_unlock(&zbpg->lock);
+ preempt_enable();
+
+ /* OK, no locks held anymore, remotify one or both zbuds */
+ if (zh1 != NULL)
+ ret = zbud_remotify_zbud(&xh1, data1, size1);
+ if (zh2 != NULL)
+ ret |= zbud_remotify_zbud(&xh2, data2, size2);
+ return ret;
+}
+
+void zbud_remotify_pages(int nr)
+{
+ struct zbud_page *zbpg;
+ int i, ret;
+
+ /*
+ * for now just try remotifying unbuddied pages, starting with
+ * least space avail
+ */
+ for (i = 0; i < MAX_CHUNK; i++) {
+retry_unbud_list_i:
+ preempt_disable(); /* enable in zbud_remotify_zbpg */
+ spin_lock_bh(&zbud_budlists_spinlock);
+ if (list_empty(&zbud_unbuddied[i].list)) {
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ preempt_enable();
+ continue; /* next i in for loop */
+ }
+ list_for_each_entry(zbpg, &zbud_unbuddied[i].list, bud_list) {
+ if (unlikely(!spin_trylock(&zbpg->lock)))
+ continue; /* next list_for_each_entry */
+ zbud_unbuddied[i].count--;
+ /* want budlists unlocked when doing zbpg remotify */
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ ret = zbud_remotify_zbpg(zbpg);
+ /* preemption is re-enabled in zbud_remotify_zbpg */
+ if (ret == 0) {
+ if (--nr <= 0)
+ goto out;
+ goto retry_unbud_list_i;
+ }
+ /* if fail to remotify any page, quit */
+ pr_err("TESTING zbud_remotify_pages failed on page,"
+ " trying to re-add\n");
+ spin_lock_bh(&zbud_budlists_spinlock);
+ spin_lock(&zbpg->lock);
+ list_add_tail(&zbpg->bud_list, &zbud_unbuddied[i].list);
+ zbud_unbuddied[i].count++;
+ spin_unlock(&zbpg->lock);
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ pr_err("TESTING zbud_remotify_pages failed on page,"
+ " finished re-add\n");
+ goto out;
+ }
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ preempt_enable();
+ }
+
+next_buddied_zbpg:
+ preempt_disable(); /* enable in zbud_remotify_zbpg */
+ spin_lock_bh(&zbud_budlists_spinlock);
+ if (list_empty(&zbud_buddied_list))
+ goto unlock_out;
+ list_for_each_entry(zbpg, &zbud_buddied_list, bud_list) {
+ if (unlikely(!spin_trylock(&zbpg->lock)))
+ continue; /* next list_for_each_entry */
+ zcache_zbud_buddied_count--;
+ /* want budlists unlocked when doing zbpg remotify */
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ ret = zbud_remotify_zbpg(zbpg);
+ /* preemption is re-enabled in zbud_remotify_zbpg */
+ if (ret == 0) {
+ if (--nr <= 0)
+ goto out;
+ goto next_buddied_zbpg;
+ }
+ /* if fail to remotify any page, quit */
+ pr_err("TESTING zbud_remotify_pages failed on BUDDIED page,"
+ " trying to re-add\n");
+ spin_lock_bh(&zbud_budlists_spinlock);
+ spin_lock(&zbpg->lock);
+ list_add_tail(&zbpg->bud_list, &zbud_buddied_list);
+ zcache_zbud_buddied_count++;
+ spin_unlock(&zbpg->lock);
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ pr_err("TESTING zbud_remotify_pages failed on BUDDIED page,"
+ " finished re-add\n");
+ goto out;
+ }
+unlock_out:
+ spin_unlock_bh(&zbud_budlists_spinlock);
+ preempt_enable();
+out:
+ return;
+}
+
+/* the "flush list" asynchronously collects pages to remotely flush */
+#define FLUSH_ENTIRE_OBJECT ((uint32_t)-1)
+static void ramster_flnode_free(struct flushlist_node *,
+ struct tmem_pool *);
+
+static void zcache_remote_flush_page(struct flushlist_node *flnode)
+{
+ struct tmem_xhandle *xh;
+ int remotenode, ret;
+
+ preempt_disable();
+ xh = &flnode->xh;
+ remotenode = flnode->xh.client_id;
+ ret = ramster_remote_flush(xh, remotenode);
+ if (ret >= 0)
+ ramster_remote_pages_flushed++;
+ else
+ ramster_remote_page_flushes_failed++;
+ preempt_enable_no_resched();
+ ramster_flnode_free(flnode, NULL);
+}
+
+static void zcache_remote_flush_object(struct flushlist_node *flnode)
+{
+ struct tmem_xhandle *xh;
+ int remotenode, ret;
+
+ preempt_disable();
+ xh = &flnode->xh;
+ remotenode = flnode->xh.client_id;
+ ret = ramster_remote_flush_object(xh, remotenode);
+ if (ret >= 0)
+ ramster_remote_objects_flushed++;
+ else
+ ramster_remote_object_flushes_failed++;
+ preempt_enable_no_resched();
+ ramster_flnode_free(flnode, NULL);
+}
+
+static void zcache_remote_eph_put(struct zbud_hdr *zbud)
+{
+ /* FIXME */
+}
+
+static void zcache_remote_pers_put(struct zv_hdr *zv)
+{
+ struct tmem_xhandle xh;
+ uint16_t size;
+ bool ephemeral;
+ int remotenode, ret = -1;
+ char *data;
+ struct tmem_pool *pool;
+ unsigned long flags;
+ unsigned char cksum;
+ char *p;
+ int i;
+ unsigned char *tmpmem = __get_cpu_var(zcache_remoteputmem);
+
+ ASSERT_SENTINEL(zv, ZVH);
+ BUG_ON(zv->client_id != LOCAL_CLIENT);
+ local_bh_disable();
+ xh.client_id = zv->client_id;
+ xh.pool_id = zv->pool_id;
+ xh.oid = zv->oid;
+ xh.index = zv->index;
+ size = xv_get_object_size(zv) - sizeof(*zv);
+ BUG_ON(size == 0 || size > zv_max_page_size);
+ data = (char *)zv + sizeof(*zv);
+ for (p = data, cksum = 0, i = 0; i < size; i++)
+ cksum += *p;
+ memcpy(tmpmem, data, size);
+ data = tmpmem;
+ pool = zcache_get_pool_by_id(zv->client_id, zv->pool_id);
+ ephemeral = is_ephemeral(pool);
+ zcache_put_pool(pool);
+ /* now OK to release lock set in caller */
+ spin_unlock(&zcache_rem_op_list_lock);
+ local_bh_enable();
+ preempt_disable();
+ ret = ramster_remote_put(&xh, data, size, ephemeral, &remotenode);
+ preempt_enable_no_resched();
+ if (ret != 0) {
+ /*
+ * This is some form of a memory leak... if the remote put
+ * fails, there will never be another attempt to remotify
+ * this page. But since we've dropped the zv pointer,
+ * the page may have been freed or the data replaced
+ * so we can't just "put it back" in the remote op list.
+ * Even if we could, not sure where to put it in the list
+ * because there may be flushes that must be strictly
+ * ordered vs the put. So leave this as a FIXME for now.
+ * But count them so we know if it becomes a problem.
+ */
+ ramster_pers_pages_remote_failed++;
+ goto out;
+ } else
+ atomic_inc(&ramster_remote_pers_pages);
+ ramster_pers_pages_remoted++;
+ /*
+ * data was successfully remoted so change the local version to
+ * point to the remote node where it landed
+ */
+ local_bh_disable();
+ pool = zcache_get_pool_by_id(LOCAL_CLIENT, xh.pool_id);
+ local_irq_save(flags);
+ (void)tmem_replace(pool, &xh.oid, xh.index,
+ pampd_make_remote(remotenode, size, cksum));
+ local_irq_restore(flags);
+ zcache_put_pool(pool);
+ local_bh_enable();
+out:
+ return;
+}
+
+static void zcache_do_remotify_ops(int nr)
+{
+ struct ramster_remotify_hdr *rem_op;
+ union remotify_list_node *u;
+
+ while (1) {
+ if (!nr)
+ goto out;
+ spin_lock(&zcache_rem_op_list_lock);
+ if (list_empty(&zcache_rem_op_list)) {
+ spin_unlock(&zcache_rem_op_list_lock);
+ goto out;
+ }
+ rem_op = list_first_entry(&zcache_rem_op_list,
+ struct ramster_remotify_hdr, list);
+ list_del_init(&rem_op->list);
+ if (rem_op->op != RAMSTER_REMOTIFY_PERS_PUT)
+ spin_unlock(&zcache_rem_op_list_lock);
+ u = (union remotify_list_node *)rem_op;
+ switch (rem_op->op) {
+ case RAMSTER_REMOTIFY_EPH_PUT:
+BUG();
+ zcache_remote_eph_put((struct zbud_hdr *)rem_op);
+ break;
+ case RAMSTER_REMOTIFY_PERS_PUT:
+ zcache_remote_pers_put((struct zv_hdr *)rem_op);
+ break;
+ case RAMSTER_REMOTIFY_FLUSH_PAGE:
+ zcache_remote_flush_page((struct flushlist_node *)u);
+ break;
+ case RAMSTER_REMOTIFY_FLUSH_OBJ:
+ zcache_remote_flush_object((struct flushlist_node *)u);
+ break;
+ default:
+ BUG();
+ }
+ }
+out:
+ return;
+}
+
+/*
+ * Communicate interface revision with userspace
+ */
+#include "cluster/ramster_nodemanager.h"
+static unsigned long ramster_interface_revision = R2NM_API_VERSION;
+
+/*
+ * For now, just push over a few pages every few seconds to
+ * ensure that it basically works
+ */
+static struct workqueue_struct *ramster_remotify_workqueue;
+static void ramster_remotify_process(struct work_struct *work);
+static DECLARE_DELAYED_WORK(ramster_remotify_worker,
+ ramster_remotify_process);
+
+static void ramster_remotify_queue_delayed_work(unsigned long delay)
+{
+ if (!queue_delayed_work(ramster_remotify_workqueue,
+ &ramster_remotify_worker, delay))
+ pr_err("ramster_remotify: bad workqueue\n");
+}
+
+
+static int use_frontswap;
+static int use_cleancache;
+static int ramster_remote_target_nodenum = -1;
+static void ramster_remotify_process(struct work_struct *work)
+{
+ static bool remotify_in_progress;
+
+ BUG_ON(irqs_disabled());
+ if (remotify_in_progress)
+ ramster_remotify_queue_delayed_work(HZ);
+ else if (ramster_remote_target_nodenum != -1) {
+ remotify_in_progress = true;
+#ifdef CONFIG_CLEANCACHE
+ if (use_cleancache && ramster_eph_remotify_enable)
+ zbud_remotify_pages(5000); /* FIXME is this a good number? */
+#endif
+#ifdef CONFIG_FRONTSWAP
+ if (use_frontswap && ramster_pers_remotify_enable)
+ zcache_do_remotify_ops(500); /* FIXME is this a good number? */
+#endif
+ remotify_in_progress = false;
+ ramster_remotify_queue_delayed_work(HZ);
+ }
+}
+
+static void ramster_remotify_init(void)
+{
+ unsigned long n = 60UL;
+ ramster_remotify_workqueue =
+ create_singlethread_workqueue("ramster_remotify");
+ ramster_remotify_queue_delayed_work(n * HZ);
+}
+
+
+static void zbud_init(void)
+{
+ int i;
+
+ INIT_LIST_HEAD(&zbud_buddied_list);
+ zcache_zbud_buddied_count = 0;
+ for (i = 0; i < NCHUNKS; i++) {
+ INIT_LIST_HEAD(&zbud_unbuddied[i].list);
+ zbud_unbuddied[i].count = 0;
+ }
+}
+
+#ifdef CONFIG_SYSFS
+/*
+ * These sysfs routines show a nice distribution of how many zbpg's are
+ * currently (and have ever been placed) in each unbuddied list. It's fun
+ * to watch but can probably go away before final merge.
+ */
+static int zbud_show_unbuddied_list_counts(char *buf)
+{
+ int i;
+ char *p = buf;
+
+ for (i = 0; i < NCHUNKS; i++)
+ p += sprintf(p, "%u ", zbud_unbuddied[i].count);
+ return p - buf;
+}
+
+static int zbud_show_cumul_chunk_counts(char *buf)
+{
+ unsigned long i, chunks = 0, total_chunks = 0, sum_total_chunks = 0;
+ unsigned long total_chunks_lte_21 = 0, total_chunks_lte_32 = 0;
+ unsigned long total_chunks_lte_42 = 0;
+ char *p = buf;
+
+ for (i = 0; i < NCHUNKS; i++) {
+ p += sprintf(p, "%lu ", zbud_cumul_chunk_counts[i]);
+ chunks += zbud_cumul_chunk_counts[i];
+ total_chunks += zbud_cumul_chunk_counts[i];
+ sum_total_chunks += i * zbud_cumul_chunk_counts[i];
+ if (i == 21)
+ total_chunks_lte_21 = total_chunks;
+ if (i == 32)
+ total_chunks_lte_32 = total_chunks;
+ if (i == 42)
+ total_chunks_lte_42 = total_chunks;
+ }
+ p += sprintf(p, "<=21:%lu <=32:%lu <=42:%lu, mean:%lu\n",
+ total_chunks_lte_21, total_chunks_lte_32, total_chunks_lte_42,
+ chunks == 0 ? 0 : sum_total_chunks / chunks);
+ return p - buf;
+}
+#endif
+
+/**********
+ * This "zv" PAM implementation combines the TLSF-based xvMalloc
+ * with lzo1x compression to maximize the amount of data that can
+ * be packed into a physical page.
+ *
+ * Zv represents a PAM page with the index and object (plus a "size" value
+ * necessary for decompression) immediately preceding the compressed data.
+ */
+
+/* rudimentary policy limits */
+/* total number of persistent pages may not exceed this percentage */
+static unsigned int zv_page_count_policy_percent = 75;
+/*
+ * byte count defining poor compression; pages with greater zsize will be
+ * rejected
+ */
+static unsigned int zv_max_zsize = (PAGE_SIZE / 8) * 7;
+/*
+ * byte count defining poor *mean* compression; pages with greater zsize
+ * will be rejected until sufficient better-compressed pages are accepted
+ * driving the mean below this threshold
+ */
+static unsigned int zv_max_mean_zsize = (PAGE_SIZE / 8) * 5;
+
+static atomic_t zv_curr_dist_counts[NCHUNKS];
+static atomic_t zv_cumul_dist_counts[NCHUNKS];
+
+
+static struct zv_hdr *zv_create(struct zcache_client *cli, uint32_t pool_id,
+ struct tmem_oid *oid, uint32_t index,
+ void *cdata, unsigned clen)
+{
+ struct page *page;
+ struct zv_hdr *zv = NULL;
+ uint32_t offset;
+ int alloc_size = clen + sizeof(struct zv_hdr);
+ int chunks = (alloc_size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
+ int ret;
+
+ BUG_ON(!irqs_disabled());
+ BUG_ON(chunks >= NCHUNKS);
+ ret = xv_malloc(cli->xvpool, clen + sizeof(struct zv_hdr),
+ &page, &offset, ZCACHE_GFP_MASK);
+ if (unlikely(ret))
+ goto out;
+ atomic_inc(&zv_curr_dist_counts[chunks]);
+ atomic_inc(&zv_cumul_dist_counts[chunks]);
- kunmap_atomic(zv, KM_USER0);
++ zv = kmap_atomic(page) + offset;
+ zv->index = index;
+ zv->oid = *oid;
+ zv->pool_id = pool_id;
+ SET_SENTINEL(zv, ZVH);
+ INIT_LIST_HEAD(&zv->rem_op.list);
+ zv->client_id = get_client_id_from_client(cli);
+ zv->rem_op.op = RAMSTER_REMOTIFY_PERS_PUT;
+ if (zv->client_id == LOCAL_CLIENT) {
+ spin_lock(&zcache_rem_op_list_lock);
+ list_add_tail(&zv->rem_op.list, &zcache_rem_op_list);
+ spin_unlock(&zcache_rem_op_list_lock);
+ }
+ memcpy((char *)zv + sizeof(struct zv_hdr), cdata, clen);
- zv = kmap_atomic(page, KM_USER0) + offset;
++ kunmap_atomic(zv);
+out:
+ return zv;
+}
+
+/* similar to zv_create, but just reserve space, no data yet */
+static struct zv_hdr *zv_alloc(struct tmem_pool *pool,
+ struct tmem_oid *oid, uint32_t index,
+ unsigned clen)
+{
+ struct zcache_client *cli = pool->client;
+ struct page *page;
+ struct zv_hdr *zv = NULL;
+ uint32_t offset;
+ int ret;
+
+ BUG_ON(!irqs_disabled());
+ BUG_ON(!is_local_client(pool->client));
+ ret = xv_malloc(cli->xvpool, clen + sizeof(struct zv_hdr),
+ &page, &offset, ZCACHE_GFP_MASK);
+ if (unlikely(ret))
+ goto out;
- kunmap_atomic(zv, KM_USER0);
++ zv = kmap_atomic(page) + offset;
+ SET_SENTINEL(zv, ZVH);
+ INIT_LIST_HEAD(&zv->rem_op.list);
+ zv->client_id = LOCAL_CLIENT;
+ zv->rem_op.op = RAMSTER_INTRANSIT_PERS;
+ zv->index = index;
+ zv->oid = *oid;
+ zv->pool_id = pool->pool_id;
- to_va = kmap_atomic(page, KM_USER0);
++ kunmap_atomic(zv);
+out:
+ return zv;
+}
+
+static void zv_free(struct xv_pool *xvpool, struct zv_hdr *zv)
+{
+ unsigned long flags;
+ struct page *page;
+ uint32_t offset;
+ uint16_t size = xv_get_object_size(zv);
+ int chunks = (size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
+
+ ASSERT_SENTINEL(zv, ZVH);
+ BUG_ON(chunks >= NCHUNKS);
+ atomic_dec(&zv_curr_dist_counts[chunks]);
+ size -= sizeof(*zv);
+ spin_lock(&zcache_rem_op_list_lock);
+ size = xv_get_object_size(zv) - sizeof(*zv);
+ BUG_ON(size == 0);
+ INVERT_SENTINEL(zv, ZVH);
+ if (!list_empty(&zv->rem_op.list))
+ list_del_init(&zv->rem_op.list);
+ spin_unlock(&zcache_rem_op_list_lock);
+ page = virt_to_page(zv);
+ offset = (unsigned long)zv & ~PAGE_MASK;
+ local_irq_save(flags);
+ xv_free(xvpool, page, offset);
+ local_irq_restore(flags);
+}
+
+static void zv_decompress(struct page *page, struct zv_hdr *zv)
+{
+ size_t clen = PAGE_SIZE;
+ char *to_va;
+ unsigned size;
+ int ret;
+
+ ASSERT_SENTINEL(zv, ZVH);
+ size = xv_get_object_size(zv) - sizeof(*zv);
+ BUG_ON(size == 0);
- kunmap_atomic(to_va, KM_USER0);
++ to_va = kmap_atomic(page);
+ ret = lzo1x_decompress_safe((char *)zv + sizeof(*zv),
+ size, to_va, &clen);
- from_va = kmap_atomic(from, KM_USER0);
++ kunmap_atomic(to_va);
+ BUG_ON(ret != LZO_E_OK);
+ BUG_ON(clen != PAGE_SIZE);
+}
+
+static void zv_copy_from_pampd(char *data, size_t *bufsize, struct zv_hdr *zv)
+{
+ unsigned size;
+
+ ASSERT_SENTINEL(zv, ZVH);
+ size = xv_get_object_size(zv) - sizeof(*zv);
+ BUG_ON(size == 0 || size > zv_max_page_size);
+ BUG_ON(size > *bufsize);
+ memcpy(data, (char *)zv + sizeof(*zv), size);
+ *bufsize = size;
+}
+
+static void zv_copy_to_pampd(struct zv_hdr *zv, char *data, size_t size)
+{
+ unsigned zv_size;
+
+ ASSERT_SENTINEL(zv, ZVH);
+ zv_size = xv_get_object_size(zv) - sizeof(*zv);
+ BUG_ON(zv_size != size);
+ BUG_ON(zv_size == 0 || zv_size > zv_max_page_size);
+ memcpy((char *)zv + sizeof(*zv), data, size);
+}
+
+#ifdef CONFIG_SYSFS
+/*
+ * show a distribution of compression stats for zv pages.
+ */
+
+static int zv_curr_dist_counts_show(char *buf)
+{
+ unsigned long i, n, chunks = 0, sum_total_chunks = 0;
+ char *p = buf;
+
+ for (i = 0; i < NCHUNKS; i++) {
+ n = atomic_read(&zv_curr_dist_counts[i]);
+ p += sprintf(p, "%lu ", n);
+ chunks += n;
+ sum_total_chunks += i * n;
+ }
+ p += sprintf(p, "mean:%lu\n",
+ chunks == 0 ? 0 : sum_total_chunks / chunks);
+ return p - buf;
+}
+
+static int zv_cumul_dist_counts_show(char *buf)
+{
+ unsigned long i, n, chunks = 0, sum_total_chunks = 0;
+ char *p = buf;
+
+ for (i = 0; i < NCHUNKS; i++) {
+ n = atomic_read(&zv_cumul_dist_counts[i]);
+ p += sprintf(p, "%lu ", n);
+ chunks += n;
+ sum_total_chunks += i * n;
+ }
+ p += sprintf(p, "mean:%lu\n",
+ chunks == 0 ? 0 : sum_total_chunks / chunks);
+ return p - buf;
+}
+
+/*
+ * setting zv_max_zsize via sysfs causes all persistent (e.g. swap)
+ * pages that don't compress to less than this value (including metadata
+ * overhead) to be rejected. We don't allow the value to get too close
+ * to PAGE_SIZE.
+ */
+static ssize_t zv_max_zsize_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *buf)
+{
+ return sprintf(buf, "%u\n", zv_max_zsize);
+}
+
+static ssize_t zv_max_zsize_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ unsigned long val;
+ int err;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ err = kstrtoul(buf, 10, &val);
+ if (err || (val == 0) || (val > (PAGE_SIZE / 8) * 7))
+ return -EINVAL;
+ zv_max_zsize = val;
+ return count;
+}
+
+/*
+ * setting zv_max_mean_zsize via sysfs causes all persistent (e.g. swap)
+ * pages that don't compress to less than this value (including metadata
+ * overhead) to be rejected UNLESS the mean compression is also smaller
+ * than this value. In other words, we are load-balancing-by-zsize the
+ * accepted pages. Again, we don't allow the value to get too close
+ * to PAGE_SIZE.
+ */
+static ssize_t zv_max_mean_zsize_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *buf)
+{
+ return sprintf(buf, "%u\n", zv_max_mean_zsize);
+}
+
+static ssize_t zv_max_mean_zsize_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ unsigned long val;
+ int err;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ err = kstrtoul(buf, 10, &val);
+ if (err || (val == 0) || (val > (PAGE_SIZE / 8) * 7))
+ return -EINVAL;
+ zv_max_mean_zsize = val;
+ return count;
+}
+
+/*
+ * setting zv_page_count_policy_percent via sysfs sets an upper bound of
+ * persistent (e.g. swap) pages that will be retained according to:
+ * (zv_page_count_policy_percent * totalram_pages) / 100)
+ * when that limit is reached, further puts will be rejected (until
+ * some pages have been flushed). Note that, due to compression,
+ * this number may exceed 100; it defaults to 75 and we set an
+ * arbitary limit of 150. A poor choice will almost certainly result
+ * in OOM's, so this value should only be changed prudently.
+ */
+static ssize_t zv_page_count_policy_percent_show(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ char *buf)
+{
+ return sprintf(buf, "%u\n", zv_page_count_policy_percent);
+}
+
+static ssize_t zv_page_count_policy_percent_store(struct kobject *kobj,
+ struct kobj_attribute *attr,
+ const char *buf, size_t count)
+{
+ unsigned long val;
+ int err;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ err = kstrtoul(buf, 10, &val);
+ if (err || (val == 0) || (val > 150))
+ return -EINVAL;
+ zv_page_count_policy_percent = val;
+ return count;
+}
+
+static struct kobj_attribute zcache_zv_max_zsize_attr = {
+ .attr = { .name = "zv_max_zsize", .mode = 0644 },
+ .show = zv_max_zsize_show,
+ .store = zv_max_zsize_store,
+};
+
+static struct kobj_attribute zcache_zv_max_mean_zsize_attr = {
+ .attr = { .name = "zv_max_mean_zsize", .mode = 0644 },
+ .show = zv_max_mean_zsize_show,
+ .store = zv_max_mean_zsize_store,
+};
+
+static struct kobj_attribute zcache_zv_page_count_policy_percent_attr = {
+ .attr = { .name = "zv_page_count_policy_percent",
+ .mode = 0644 },
+ .show = zv_page_count_policy_percent_show,
+ .store = zv_page_count_policy_percent_store,
+};
+#endif
+
+/*
+ * zcache core code starts here
+ */
+
+/* useful stats not collected by cleancache or frontswap */
+static unsigned long zcache_flush_total;
+static unsigned long zcache_flush_found;
+static unsigned long zcache_flobj_total;
+static unsigned long zcache_flobj_found;
+static unsigned long zcache_failed_eph_puts;
+static unsigned long zcache_nonactive_puts;
+static unsigned long zcache_failed_pers_puts;
+
+/*
+ * Tmem operations assume the poolid implies the invoking client.
+ * Zcache only has one client (the kernel itself): LOCAL_CLIENT.
+ * RAMster has each client numbered by cluster node, and a KVM version
+ * of zcache would have one client per guest and each client might
+ * have a poolid==N.
+ */
+static struct tmem_pool *zcache_get_pool_by_id(uint16_t cli_id, uint16_t poolid)
+{
+ struct tmem_pool *pool = NULL;
+ struct zcache_client *cli = NULL;
+
+ if (cli_id == LOCAL_CLIENT)
+ cli = &zcache_host;
+ else {
+ if (cli_id >= MAX_CLIENTS)
+ goto out;
+ cli = &zcache_clients[cli_id];
+ if (cli == NULL)
+ goto out;
+ atomic_inc(&cli->refcount);
+ }
+ if (poolid < MAX_POOLS_PER_CLIENT) {
+ pool = cli->tmem_pools[poolid];
+ if (pool != NULL)
+ atomic_inc(&pool->refcount);
+ }
+out:
+ return pool;
+}
+
+static void zcache_put_pool(struct tmem_pool *pool)
+{
+ struct zcache_client *cli = NULL;
+
+ if (pool == NULL)
+ BUG();
+ cli = pool->client;
+ atomic_dec(&pool->refcount);
+ atomic_dec(&cli->refcount);
+}
+
+int zcache_new_client(uint16_t cli_id)
+{
+ struct zcache_client *cli = NULL;
+ int ret = -1;
+
+ if (cli_id == LOCAL_CLIENT)
+ cli = &zcache_host;
+ else if ((unsigned int)cli_id < MAX_CLIENTS)
+ cli = &zcache_clients[cli_id];
+ if (cli == NULL)
+ goto out;
+ if (cli->allocated)
+ goto out;
+ cli->allocated = 1;
+#ifdef CONFIG_FRONTSWAP
+ cli->xvpool = xv_create_pool();
+ if (cli->xvpool == NULL)
+ goto out;
+#endif
+ ret = 0;
+out:
+ return ret;
+}
+
+/* counters for debugging */
+static unsigned long zcache_failed_get_free_pages;
+static unsigned long zcache_failed_alloc;
+static unsigned long zcache_put_to_flush;
+
+/*
+ * for now, used named slabs so can easily track usage; later can
+ * either just use kmalloc, or perhaps add a slab-like allocator
+ * to more carefully manage total memory utilization
+ */
+static struct kmem_cache *zcache_objnode_cache;
+static struct kmem_cache *zcache_obj_cache;
+static struct kmem_cache *ramster_flnode_cache;
+static atomic_t zcache_curr_obj_count = ATOMIC_INIT(0);
+static unsigned long zcache_curr_obj_count_max;
+static atomic_t zcache_curr_objnode_count = ATOMIC_INIT(0);
+static unsigned long zcache_curr_objnode_count_max;
+
+/*
+ * to avoid memory allocation recursion (e.g. due to direct reclaim), we
+ * preload all necessary data structures so the hostops callbacks never
+ * actually do a malloc
+ */
+struct zcache_preload {
+ void *page;
+ struct tmem_obj *obj;
+ int nr;
+ struct tmem_objnode *objnodes[OBJNODE_TREE_MAX_PATH];
+ struct flushlist_node *flnode;
+};
+static DEFINE_PER_CPU(struct zcache_preload, zcache_preloads) = { 0, };
+
+static int zcache_do_preload(struct tmem_pool *pool)
+{
+ struct zcache_preload *kp;
+ struct tmem_objnode *objnode;
+ struct tmem_obj *obj;
+ struct flushlist_node *flnode;
+ void *page;
+ int ret = -ENOMEM;
+
+ if (unlikely(zcache_objnode_cache == NULL))
+ goto out;
+ if (unlikely(zcache_obj_cache == NULL))
+ goto out;
+ preempt_disable();
+ kp = &__get_cpu_var(zcache_preloads);
+ while (kp->nr < ARRAY_SIZE(kp->objnodes)) {
+ preempt_enable_no_resched();
+ objnode = kmem_cache_alloc(zcache_objnode_cache,
+ ZCACHE_GFP_MASK);
+ if (unlikely(objnode == NULL)) {
+ zcache_failed_alloc++;
+ goto out;
+ }
+ preempt_disable();
+ kp = &__get_cpu_var(zcache_preloads);
+ if (kp->nr < ARRAY_SIZE(kp->objnodes))
+ kp->objnodes[kp->nr++] = objnode;
+ else
+ kmem_cache_free(zcache_objnode_cache, objnode);
+ }
+ preempt_enable_no_resched();
+ obj = kmem_cache_alloc(zcache_obj_cache, ZCACHE_GFP_MASK);
+ if (unlikely(obj == NULL)) {
+ zcache_failed_alloc++;
+ goto out;
+ }
+ flnode = kmem_cache_alloc(ramster_flnode_cache, ZCACHE_GFP_MASK);
+ if (unlikely(flnode == NULL)) {
+ zcache_failed_alloc++;
+ goto out;
+ }
+ if (is_ephemeral(pool)) {
+ page = (void *)__get_free_page(ZCACHE_GFP_MASK);
+ if (unlikely(page == NULL)) {
+ zcache_failed_get_free_pages++;
+ kmem_cache_free(zcache_obj_cache, obj);
+ kmem_cache_free(ramster_flnode_cache, flnode);
+ goto out;
+ }
+ }
+ preempt_disable();
+ kp = &__get_cpu_var(zcache_preloads);
+ if (kp->obj == NULL)
+ kp->obj = obj;
+ else
+ kmem_cache_free(zcache_obj_cache, obj);
+ if (kp->flnode == NULL)
+ kp->flnode = flnode;
+ else
+ kmem_cache_free(ramster_flnode_cache, flnode);
+ if (is_ephemeral(pool)) {
+ if (kp->page == NULL)
+ kp->page = page;
+ else
+ free_page((unsigned long)page);
+ }
+ ret = 0;
+out:
+ return ret;
+}
+
+static int ramster_do_preload_flnode_only(struct tmem_pool *pool)
+{
+ struct zcache_preload *kp;
+ struct flushlist_node *flnode;
+ int ret = -ENOMEM;
+
+ BUG_ON(!irqs_disabled());
+ if (unlikely(ramster_flnode_cache == NULL))
+ BUG();
+ kp = &__get_cpu_var(zcache_preloads);
+ flnode = kmem_cache_alloc(ramster_flnode_cache, GFP_ATOMIC);
+ if (unlikely(flnode == NULL) && kp->flnode == NULL)
+ BUG(); /* FIXME handle more gracefully, but how??? */
+ else if (kp->flnode == NULL)
+ kp->flnode = flnode;
+ else
+ kmem_cache_free(ramster_flnode_cache, flnode);
+ return ret;
+}
+
+static void *zcache_get_free_page(void)
+{
+ struct zcache_preload *kp;
+ void *page;
+
+ kp = &__get_cpu_var(zcache_preloads);
+ page = kp->page;
+ BUG_ON(page == NULL);
+ kp->page = NULL;
+ return page;
+}
+
+static void zcache_free_page(void *p)
+{
+ free_page((unsigned long)p);
+}
+
+/*
+ * zcache implementation for tmem host ops
+ */
+
+static struct tmem_objnode *zcache_objnode_alloc(struct tmem_pool *pool)
+{
+ struct tmem_objnode *objnode = NULL;
+ unsigned long count;
+ struct zcache_preload *kp;
+
+ kp = &__get_cpu_var(zcache_preloads);
+ if (kp->nr <= 0)
+ goto out;
+ objnode = kp->objnodes[kp->nr - 1];
+ BUG_ON(objnode == NULL);
+ kp->objnodes[kp->nr - 1] = NULL;
+ kp->nr--;
+ count = atomic_inc_return(&zcache_curr_objnode_count);
+ if (count > zcache_curr_objnode_count_max)
+ zcache_curr_objnode_count_max = count;
+out:
+ return objnode;
+}
+
+static void zcache_objnode_free(struct tmem_objnode *objnode,
+ struct tmem_pool *pool)
+{
+ atomic_dec(&zcache_curr_objnode_count);
+ BUG_ON(atomic_read(&zcache_curr_objnode_count) < 0);
+ kmem_cache_free(zcache_objnode_cache, objnode);
+}
+
+static struct tmem_obj *zcache_obj_alloc(struct tmem_pool *pool)
+{
+ struct tmem_obj *obj = NULL;
+ unsigned long count;
+ struct zcache_preload *kp;
+
+ kp = &__get_cpu_var(zcache_preloads);
+ obj = kp->obj;
+ BUG_ON(obj == NULL);
+ kp->obj = NULL;
+ count = atomic_inc_return(&zcache_curr_obj_count);
+ if (count > zcache_curr_obj_count_max)
+ zcache_curr_obj_count_max = count;
+ return obj;
+}
+
+static void zcache_obj_free(struct tmem_obj *obj, struct tmem_pool *pool)
+{
+ atomic_dec(&zcache_curr_obj_count);
+ BUG_ON(atomic_read(&zcache_curr_obj_count) < 0);
+ kmem_cache_free(zcache_obj_cache, obj);
+}
+
+static struct flushlist_node *ramster_flnode_alloc(struct tmem_pool *pool)
+{
+ struct flushlist_node *flnode = NULL;
+ struct zcache_preload *kp;
+ int count;
+
+ kp = &__get_cpu_var(zcache_preloads);
+ flnode = kp->flnode;
+ BUG_ON(flnode == NULL);
+ kp->flnode = NULL;
+ count = atomic_inc_return(&ramster_curr_flnode_count);
+ if (count > ramster_curr_flnode_count_max)
+ ramster_curr_flnode_count_max = count;
+ return flnode;
+}
+
+static void ramster_flnode_free(struct flushlist_node *flnode,
+ struct tmem_pool *pool)
+{
+ atomic_dec(&ramster_curr_flnode_count);
+ BUG_ON(atomic_read(&ramster_curr_flnode_count) < 0);
+ kmem_cache_free(ramster_flnode_cache, flnode);
+}
+
+static struct tmem_hostops zcache_hostops = {
+ .obj_alloc = zcache_obj_alloc,
+ .obj_free = zcache_obj_free,
+ .objnode_alloc = zcache_objnode_alloc,
+ .objnode_free = zcache_objnode_free,
+};
+
+/*
+ * zcache implementations for PAM page descriptor ops
+ */
+
+
+static inline void dec_and_check(atomic_t *pvar)
+{
+ atomic_dec(pvar);
+ /* later when all accounting is fixed, make this a BUG */
+ WARN_ON_ONCE(atomic_read(pvar) < 0);
+}
+
+static atomic_t zcache_curr_eph_pampd_count = ATOMIC_INIT(0);
+static unsigned long zcache_curr_eph_pampd_count_max;
+static atomic_t zcache_curr_pers_pampd_count = ATOMIC_INIT(0);
+static unsigned long zcache_curr_pers_pampd_count_max;
+
+/* forward reference */
+static int zcache_compress(struct page *from, void **out_va, size_t *out_len);
+
+static int zcache_pampd_eph_create(char *data, size_t size, bool raw,
+ struct tmem_pool *pool, struct tmem_oid *oid,
+ uint32_t index, void **pampd)
+{
+ int ret = -1;
+ void *cdata = data;
+ size_t clen = size;
+ struct zcache_client *cli = pool->client;
+ uint16_t client_id = get_client_id_from_client(cli);
+ struct page *page = NULL;
+ unsigned long count;
+
+ if (!raw) {
+ page = virt_to_page(data);
+ ret = zcache_compress(page, &cdata, &clen);
+ if (ret == 0)
+ goto out;
+ if (clen == 0 || clen > zbud_max_buddy_size()) {
+ zcache_compress_poor++;
+ goto out;
+ }
+ }
+ *pampd = (void *)zbud_create(client_id, pool->pool_id, oid,
+ index, page, cdata, clen);
+ if (*pampd == NULL) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ ret = 0;
+ count = atomic_inc_return(&zcache_curr_eph_pampd_count);
+ if (count > zcache_curr_eph_pampd_count_max)
+ zcache_curr_eph_pampd_count_max = count;
+ if (client_id != LOCAL_CLIENT) {
+ count = atomic_inc_return(&ramster_foreign_eph_pampd_count);
+ if (count > ramster_foreign_eph_pampd_count_max)
+ ramster_foreign_eph_pampd_count_max = count;
+ }
+out:
+ return ret;
+}
+
+static int zcache_pampd_pers_create(char *data, size_t size, bool raw,
+ struct tmem_pool *pool, struct tmem_oid *oid,
+ uint32_t index, void **pampd)
+{
+ int ret = -1;
+ void *cdata = data;
+ size_t clen = size;
+ struct zcache_client *cli = pool->client;
+ struct page *page;
+ unsigned long count;
+ unsigned long zv_mean_zsize;
+ struct zv_hdr *zv;
+ long curr_pers_pampd_count;
+ u64 total_zsize;
+#ifdef RAMSTER_TESTING
+ static bool pampd_neg_warned;
+#endif
+
+ curr_pers_pampd_count = atomic_read(&zcache_curr_pers_pampd_count) -
+ atomic_read(&ramster_remote_pers_pages);
+#ifdef RAMSTER_TESTING
+ /* should always be positive, but warn if accounting is off */
+ if (!pampd_neg_warned) {
+ pr_warn("ramster: bad accounting for curr_pers_pampd_count\n");
+ pampd_neg_warned = true;
+ }
+#endif
+ if (curr_pers_pampd_count >
+ (zv_page_count_policy_percent * totalram_pages) / 100) {
+ zcache_policy_percent_exceeded++;
+ goto out;
+ }
+ if (raw)
+ goto ok_to_create;
+ page = virt_to_page(data);
+ if (zcache_compress(page, &cdata, &clen) == 0)
+ goto out;
+ /* reject if compression is too poor */
+ if (clen > zv_max_zsize) {
+ zcache_compress_poor++;
+ goto out;
+ }
+ /* reject if mean compression is too poor */
+ if ((clen > zv_max_mean_zsize) && (curr_pers_pampd_count > 0)) {
+ total_zsize = xv_get_total_size_bytes(cli->xvpool);
+ zv_mean_zsize = div_u64(total_zsize, curr_pers_pampd_count);
+ if (zv_mean_zsize > zv_max_mean_zsize) {
+ zcache_mean_compress_poor++;
+ goto out;
+ }
+ }
+ok_to_create:
+ *pampd = (void *)zv_create(cli, pool->pool_id, oid, index, cdata, clen);
+ if (*pampd == NULL) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ ret = 0;
+ count = atomic_inc_return(&zcache_curr_pers_pampd_count);
+ if (count > zcache_curr_pers_pampd_count_max)
+ zcache_curr_pers_pampd_count_max = count;
+ if (is_local_client(cli))
+ goto out;
+ zv = *(struct zv_hdr **)pampd;
+ count = atomic_inc_return(&ramster_foreign_pers_pampd_count);
+ if (count > ramster_foreign_pers_pampd_count_max)
+ ramster_foreign_pers_pampd_count_max = count;
+out:
+ return ret;
+}
+
+static void *zcache_pampd_create(char *data, size_t size, bool raw, int eph,
+ struct tmem_pool *pool, struct tmem_oid *oid,
+ uint32_t index)
+{
+ void *pampd = NULL;
+ int ret;
+ bool ephemeral;
+
+ BUG_ON(preemptible());
+ ephemeral = (eph == 1) || ((eph == 0) && is_ephemeral(pool));
+ if (ephemeral)
+ ret = zcache_pampd_eph_create(data, size, raw, pool,
+ oid, index, &pampd);
+ else
+ ret = zcache_pampd_pers_create(data, size, raw, pool,
+ oid, index, &pampd);
+ /* FIXME add some counters here for failed creates? */
+ return pampd;
+}
+
+/*
+ * fill the pageframe corresponding to the struct page with the data
+ * from the passed pampd
+ */
+static int zcache_pampd_get_data(char *data, size_t *bufsize, bool raw,
+ void *pampd, struct tmem_pool *pool,
+ struct tmem_oid *oid, uint32_t index)
+{
+ int ret = 0;
+
+ BUG_ON(preemptible());
+ BUG_ON(is_ephemeral(pool)); /* Fix later for shared pools? */
+ BUG_ON(pampd_is_remote(pampd));
+ if (raw)
+ zv_copy_from_pampd(data, bufsize, pampd);
+ else
+ zv_decompress(virt_to_page(data), pampd);
+ return ret;
+}
+
+static int zcache_pampd_get_data_and_free(char *data, size_t *bufsize, bool raw,
+ void *pampd, struct tmem_pool *pool,
+ struct tmem_oid *oid, uint32_t index)
+{
+ int ret = 0;
+ unsigned long flags;
+ struct zcache_client *cli = pool->client;
+
+ BUG_ON(preemptible());
+ BUG_ON(pampd_is_remote(pampd));
+ if (is_ephemeral(pool)) {
+ local_irq_save(flags);
+ if (raw)
+ zbud_copy_from_pampd(data, bufsize, pampd);
+ else
+ ret = zbud_decompress(virt_to_page(data), pampd);
+ zbud_free_and_delist((struct zbud_hdr *)pampd);
+ local_irq_restore(flags);
+ if (!is_local_client(cli))
+ dec_and_check(&ramster_foreign_eph_pampd_count);
+ dec_and_check(&zcache_curr_eph_pampd_count);
+ } else {
+ if (is_local_client(cli))
+ BUG();
+ if (raw)
+ zv_copy_from_pampd(data, bufsize, pampd);
+ else
+ zv_decompress(virt_to_page(data), pampd);
+ zv_free(cli->xvpool, pampd);
+ if (!is_local_client(cli))
+ dec_and_check(&ramster_foreign_pers_pampd_count);
+ dec_and_check(&zcache_curr_pers_pampd_count);
+ ret = 0;
+ }
+ return ret;
+}
+
+static bool zcache_pampd_is_remote(void *pampd)
+{
+ return pampd_is_remote(pampd);
+}
+
+/*
+ * free the pampd and remove it from any zcache lists
+ * pampd must no longer be pointed to from any tmem data structures!
+ */
+static void zcache_pampd_free(void *pampd, struct tmem_pool *pool,
+ struct tmem_oid *oid, uint32_t index, bool acct)
+{
+ struct zcache_client *cli = pool->client;
+ bool eph = is_ephemeral(pool);
+ struct zv_hdr *zv;
+
+ BUG_ON(preemptible());
+ if (pampd_is_remote(pampd)) {
+ WARN_ON(acct == false);
+ if (oid == NULL) {
+ /*
+ * a NULL oid means to ignore this pampd free
+ * as the remote freeing will be handled elsewhere
+ */
+ } else if (eph) {
+ /* FIXME remote flush optional but probably good idea */
+ /* FIXME get these working properly again */
+ dec_and_check(&zcache_curr_eph_pampd_count);
+ } else if (pampd_is_intransit(pampd)) {
+ /* did a pers remote get_and_free, so just free local */
+ pampd = pampd_mask_intransit_and_remote(pampd);
+ goto local_pers;
+ } else {
+ struct flushlist_node *flnode =
+ ramster_flnode_alloc(pool);
+
+ flnode->xh.client_id = pampd_remote_node(pampd);
+ flnode->xh.pool_id = pool->pool_id;
+ flnode->xh.oid = *oid;
+ flnode->xh.index = index;
+ flnode->rem_op.op = RAMSTER_REMOTIFY_FLUSH_PAGE;
+ spin_lock(&zcache_rem_op_list_lock);
+ list_add(&flnode->rem_op.list, &zcache_rem_op_list);
+ spin_unlock(&zcache_rem_op_list_lock);
+ dec_and_check(&zcache_curr_pers_pampd_count);
+ dec_and_check(&ramster_remote_pers_pages);
+ }
+ } else if (eph) {
+ zbud_free_and_delist((struct zbud_hdr *)pampd);
+ if (!is_local_client(pool->client))
+ dec_and_check(&ramster_foreign_eph_pampd_count);
+ if (acct)
+ /* FIXME get these working properly again */
+ dec_and_check(&zcache_curr_eph_pampd_count);
+ } else {
+local_pers:
+ zv = (struct zv_hdr *)pampd;
+ if (!is_local_client(pool->client))
+ dec_and_check(&ramster_foreign_pers_pampd_count);
+ zv_free(cli->xvpool, zv);
+ if (acct)
+ /* FIXME get these working properly again */
+ dec_and_check(&zcache_curr_pers_pampd_count);
+ }
+}
+
+static void zcache_pampd_free_obj(struct tmem_pool *pool,
+ struct tmem_obj *obj)
+{
+ struct flushlist_node *flnode;
+
+ BUG_ON(preemptible());
+ if (obj->extra == NULL)
+ return;
+ BUG_ON(!pampd_is_remote(obj->extra));
+ flnode = ramster_flnode_alloc(pool);
+ flnode->xh.client_id = pampd_remote_node(obj->extra);
+ flnode->xh.pool_id = pool->pool_id;
+ flnode->xh.oid = obj->oid;
+ flnode->xh.index = FLUSH_ENTIRE_OBJECT;
+ flnode->rem_op.op = RAMSTER_REMOTIFY_FLUSH_OBJ;
+ spin_lock(&zcache_rem_op_list_lock);
+ list_add(&flnode->rem_op.list, &zcache_rem_op_list);
+ spin_unlock(&zcache_rem_op_list_lock);
+}
+
+void zcache_pampd_new_obj(struct tmem_obj *obj)
+{
+ obj->extra = NULL;
+}
+
+int zcache_pampd_replace_in_obj(void *new_pampd, struct tmem_obj *obj)
+{
+ int ret = -1;
+
+ if (new_pampd != NULL) {
+ if (obj->extra == NULL)
+ obj->extra = new_pampd;
+ /* enforce that all remote pages in an object reside
+ * in the same node! */
+ else if (pampd_remote_node(new_pampd) !=
+ pampd_remote_node((void *)(obj->extra)))
+ BUG();
+ ret = 0;
+ }
+ return ret;
+}
+
+/*
+ * Called by the message handler after a (still compressed) page has been
+ * fetched from the remote machine in response to an "is_remote" tmem_get
+ * or persistent tmem_localify. For a tmem_get, "extra" is the address of
+ * the page that is to be filled to succesfully resolve the tmem_get; for
+ * a (persistent) tmem_localify, "extra" is NULL (as the data is placed only
+ * in the local zcache). "data" points to "size" bytes of (compressed) data
+ * passed in the message. In the case of a persistent remote get, if
+ * pre-allocation was successful (see zcache_repatriate_preload), the page
+ * is placed into both local zcache and at "extra".
+ */
+int zcache_localify(int pool_id, struct tmem_oid *oidp,
+ uint32_t index, char *data, size_t size,
+ void *extra)
+{
+ int ret = -ENOENT;
+ unsigned long flags;
+ struct tmem_pool *pool;
+ bool ephemeral, delete = false;
+ size_t clen = PAGE_SIZE;
+ void *pampd, *saved_hb;
+ struct tmem_obj *obj;
+
+ pool = zcache_get_pool_by_id(LOCAL_CLIENT, pool_id);
+ if (unlikely(pool == NULL))
+ /* pool doesn't exist anymore */
+ goto out;
+ ephemeral = is_ephemeral(pool);
+ local_irq_save(flags); /* FIXME: maybe only disable softirqs? */
+ pampd = tmem_localify_get_pampd(pool, oidp, index, &obj, &saved_hb);
+ if (pampd == NULL) {
+ /* hmmm... must have been a flush while waiting */
+#ifdef RAMSTER_TESTING
+ pr_err("UNTESTED pampd==NULL in zcache_localify\n");
+#endif
+ if (ephemeral)
+ ramster_remote_eph_pages_unsucc_get++;
+ else
+ ramster_remote_pers_pages_unsucc_get++;
+ obj = NULL;
+ goto finish;
+ } else if (unlikely(!pampd_is_remote(pampd))) {
+ /* hmmm... must have been a dup put while waiting */
+#ifdef RAMSTER_TESTING
+ pr_err("UNTESTED dup while waiting in zcache_localify\n");
+#endif
+ if (ephemeral)
+ ramster_remote_eph_pages_unsucc_get++;
+ else
+ ramster_remote_pers_pages_unsucc_get++;
+ obj = NULL;
+ pampd = NULL;
+ ret = -EEXIST;
+ goto finish;
+ } else if (size == 0) {
+ /* no remote data, delete the local is_remote pampd */
+ pampd = NULL;
+ if (ephemeral)
+ ramster_remote_eph_pages_unsucc_get++;
+ else
+ BUG();
+ delete = true;
+ goto finish;
+ }
+ if (!ephemeral && pampd_is_intransit(pampd)) {
+ /* localify to zcache */
+ pampd = pampd_mask_intransit_and_remote(pampd);
+ zv_copy_to_pampd(pampd, data, size);
+ } else {
+ pampd = NULL;
+ obj = NULL;
+ }
+ if (extra != NULL) {
+ /* decompress direct-to-memory to complete remotify */
+ ret = lzo1x_decompress_safe((char *)data, size,
+ (char *)extra, &clen);
+ BUG_ON(ret != LZO_E_OK);
+ BUG_ON(clen != PAGE_SIZE);
+ }
+ if (ephemeral)
+ ramster_remote_eph_pages_succ_get++;
+ else
+ ramster_remote_pers_pages_succ_get++;
+ ret = 0;
+finish:
+ tmem_localify_finish(obj, index, pampd, saved_hb, delete);
+ zcache_put_pool(pool);
+ local_irq_restore(flags);
+out:
+ return ret;
+}
+
+/*
+ * Called on a remote persistent tmem_get to attempt to preallocate
+ * local storage for the data contained in the remote persistent page.
+ * If succesfully preallocated, returns the pampd, marked as remote and
+ * in_transit. Else returns NULL. Note that the appropriate tmem data
+ * structure must be locked.
+ */
+static void *zcache_pampd_repatriate_preload(void *pampd,
+ struct tmem_pool *pool,
+ struct tmem_oid *oid,
+ uint32_t index,
+ bool *intransit)
+{
+ int clen = pampd_remote_size(pampd);
+ void *ret_pampd = NULL;
+ unsigned long flags;
+
+ if (!pampd_is_remote(pampd))
+ BUG();
+ if (is_ephemeral(pool))
+ BUG();
+ if (pampd_is_intransit(pampd)) {
+ /*
+ * to avoid multiple allocations (and maybe a memory leak)
+ * don't preallocate if already in the process of being
+ * repatriated
+ */
+ *intransit = true;
+ goto out;
+ }
+ *intransit = false;
+ local_irq_save(flags);
+ ret_pampd = (void *)zv_alloc(pool, oid, index, clen);
+ if (ret_pampd != NULL) {
+ /*
+ * a pampd is marked intransit if it is remote and space has
+ * been allocated for it locally (note, only happens for
+ * persistent pages, in which case the remote copy is freed)
+ */
+ ret_pampd = pampd_mark_intransit(ret_pampd);
+ dec_and_check(&ramster_remote_pers_pages);
+ } else
+ ramster_pers_pages_remote_nomem++;
+ local_irq_restore(flags);
+out:
+ return ret_pampd;
+}
+
+/*
+ * Called on a remote tmem_get to invoke a message to fetch the page.
+ * Might sleep so no tmem locks can be held. "extra" is passed
+ * all the way through the round-trip messaging to zcache_localify.
+ */
+static int zcache_pampd_repatriate(void *fake_pampd, void *real_pampd,
+ struct tmem_pool *pool,
+ struct tmem_oid *oid, uint32_t index,
+ bool free, void *extra)
+{
+ struct tmem_xhandle xh;
+ int ret;
+
+ if (pampd_is_intransit(real_pampd))
+ /* have local space pre-reserved, so free remote copy */
+ free = true;
+ xh = tmem_xhandle_fill(LOCAL_CLIENT, pool, oid, index);
+ /* unreliable request/response for now */
+ ret = ramster_remote_async_get(&xh, free,
+ pampd_remote_node(fake_pampd),
+ pampd_remote_size(fake_pampd),
+ pampd_remote_cksum(fake_pampd),
+ extra);
+#ifdef RAMSTER_TESTING
+ if (ret != 0 && ret != -ENOENT)
+ pr_err("TESTING zcache_pampd_repatriate returns, ret=%d\n",
+ ret);
+#endif
+ return ret;
+}
+
+static struct tmem_pamops zcache_pamops = {
+ .create = zcache_pampd_create,
+ .get_data = zcache_pampd_get_data,
+ .free = zcache_pampd_free,
+ .get_data_and_free = zcache_pampd_get_data_and_free,
+ .free_obj = zcache_pampd_free_obj,
+ .is_remote = zcache_pampd_is_remote,
+ .repatriate_preload = zcache_pampd_repatriate_preload,
+ .repatriate = zcache_pampd_repatriate,
+ .new_obj = zcache_pampd_new_obj,
+ .replace_in_obj = zcache_pampd_replace_in_obj,
+};
+
+/*
+ * zcache compression/decompression and related per-cpu stuff
+ */
+
+#define LZO_WORKMEM_BYTES LZO1X_1_MEM_COMPRESS
+#define LZO_DSTMEM_PAGE_ORDER 1
+static DEFINE_PER_CPU(unsigned char *, zcache_workmem);
+static DEFINE_PER_CPU(unsigned char *, zcache_dstmem);
+
+static int zcache_compress(struct page *from, void **out_va, size_t *out_len)
+{
+ int ret = 0;
+ unsigned char *dmem = __get_cpu_var(zcache_dstmem);
+ unsigned char *wmem = __get_cpu_var(zcache_workmem);
+ char *from_va;
+
+ BUG_ON(!irqs_disabled());
+ if (unlikely(dmem == NULL || wmem == NULL))
+ goto out; /* no buffer, so can't compress */
- kunmap_atomic(from_va, KM_USER0);
++ from_va = kmap_atomic(from);
+ mb();
+ ret = lzo1x_1_compress(from_va, PAGE_SIZE, dmem, out_len, wmem);
+ BUG_ON(ret != LZO_E_OK);
+ *out_va = dmem;
++ kunmap_atomic(from_va);
+ ret = 1;
+out:
+ return ret;
+}
+
+
+static int zcache_cpu_notifier(struct notifier_block *nb,
+ unsigned long action, void *pcpu)
+{
+ int cpu = (long)pcpu;
+ struct zcache_preload *kp;
+
+ switch (action) {
+ case CPU_UP_PREPARE:
+ per_cpu(zcache_dstmem, cpu) = (void *)__get_free_pages(
+ GFP_KERNEL | __GFP_REPEAT,
+ LZO_DSTMEM_PAGE_ORDER),
+ per_cpu(zcache_workmem, cpu) =
+ kzalloc(LZO1X_MEM_COMPRESS,
+ GFP_KERNEL | __GFP_REPEAT);
+ per_cpu(zcache_remoteputmem, cpu) =
+ kzalloc(PAGE_SIZE, GFP_KERNEL | __GFP_REPEAT);
+ break;
+ case CPU_DEAD:
+ case CPU_UP_CANCELED:
+ kfree(per_cpu(zcache_remoteputmem, cpu));
+ per_cpu(zcache_remoteputmem, cpu) = NULL;
+ free_pages((unsigned long)per_cpu(zcache_dstmem, cpu),
+ LZO_DSTMEM_PAGE_ORDER);
+ per_cpu(zcache_dstmem, cpu) = NULL;
+ kfree(per_cpu(zcache_workmem, cpu));
+ per_cpu(zcache_workmem, cpu) = NULL;
+ kp = &per_cpu(zcache_preloads, cpu);
+ while (kp->nr) {
+ kmem_cache_free(zcache_objnode_cache,
+ kp->objnodes[kp->nr - 1]);
+ kp->objnodes[kp->nr - 1] = NULL;
+ kp->nr--;
+ }
+ if (kp->obj) {
+ kmem_cache_free(zcache_obj_cache, kp->obj);
+ kp->obj = NULL;
+ }
+ if (kp->flnode) {
+ kmem_cache_free(ramster_flnode_cache, kp->flnode);
+ kp->flnode = NULL;
+ }
+ if (kp->page) {
+ free_page((unsigned long)kp->page);
+ kp->page = NULL;
+ }
+ break;
+ default:
+ break;
+ }
+ return NOTIFY_OK;
+}
+
+static struct notifier_block zcache_cpu_notifier_block = {
+ .notifier_call = zcache_cpu_notifier
+};
+
+#ifdef CONFIG_SYSFS
+#define ZCACHE_SYSFS_RO(_name) \
+ static ssize_t zcache_##_name##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, char *buf) \
+ { \
+ return sprintf(buf, "%lu\n", zcache_##_name); \
+ } \
+ static struct kobj_attribute zcache_##_name##_attr = { \
+ .attr = { .name = __stringify(_name), .mode = 0444 }, \
+ .show = zcache_##_name##_show, \
+ }
+
+#define ZCACHE_SYSFS_RO_ATOMIC(_name) \
+ static ssize_t zcache_##_name##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, char *buf) \
+ { \
+ return sprintf(buf, "%d\n", atomic_read(&zcache_##_name)); \
+ } \
+ static struct kobj_attribute zcache_##_name##_attr = { \
+ .attr = { .name = __stringify(_name), .mode = 0444 }, \
+ .show = zcache_##_name##_show, \
+ }
+
+#define ZCACHE_SYSFS_RO_CUSTOM(_name, _func) \
+ static ssize_t zcache_##_name##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, char *buf) \
+ { \
+ return _func(buf); \
+ } \
+ static struct kobj_attribute zcache_##_name##_attr = { \
+ .attr = { .name = __stringify(_name), .mode = 0444 }, \
+ .show = zcache_##_name##_show, \
+ }
+
+ZCACHE_SYSFS_RO(curr_obj_count_max);
+ZCACHE_SYSFS_RO(curr_objnode_count_max);
+ZCACHE_SYSFS_RO(flush_total);
+ZCACHE_SYSFS_RO(flush_found);
+ZCACHE_SYSFS_RO(flobj_total);
+ZCACHE_SYSFS_RO(flobj_found);
+ZCACHE_SYSFS_RO(failed_eph_puts);
+ZCACHE_SYSFS_RO(nonactive_puts);
+ZCACHE_SYSFS_RO(failed_pers_puts);
+ZCACHE_SYSFS_RO(zbud_curr_zbytes);
+ZCACHE_SYSFS_RO(zbud_cumul_zpages);
+ZCACHE_SYSFS_RO(zbud_cumul_zbytes);
+ZCACHE_SYSFS_RO(zbud_buddied_count);
+ZCACHE_SYSFS_RO(evicted_raw_pages);
+ZCACHE_SYSFS_RO(evicted_unbuddied_pages);
+ZCACHE_SYSFS_RO(evicted_buddied_pages);
+ZCACHE_SYSFS_RO(failed_get_free_pages);
+ZCACHE_SYSFS_RO(failed_alloc);
+ZCACHE_SYSFS_RO(put_to_flush);
+ZCACHE_SYSFS_RO(compress_poor);
+ZCACHE_SYSFS_RO(mean_compress_poor);
+ZCACHE_SYSFS_RO(policy_percent_exceeded);
+ZCACHE_SYSFS_RO_ATOMIC(zbud_curr_raw_pages);
+ZCACHE_SYSFS_RO_ATOMIC(zbud_curr_zpages);
+ZCACHE_SYSFS_RO_ATOMIC(curr_obj_count);
+ZCACHE_SYSFS_RO_ATOMIC(curr_objnode_count);
+ZCACHE_SYSFS_RO_CUSTOM(zbud_unbuddied_list_counts,
+ zbud_show_unbuddied_list_counts);
+ZCACHE_SYSFS_RO_CUSTOM(zbud_cumul_chunk_counts,
+ zbud_show_cumul_chunk_counts);
+ZCACHE_SYSFS_RO_CUSTOM(zv_curr_dist_counts,
+ zv_curr_dist_counts_show);
+ZCACHE_SYSFS_RO_CUSTOM(zv_cumul_dist_counts,
+ zv_cumul_dist_counts_show);
+
+static struct attribute *zcache_attrs[] = {
+ &zcache_curr_obj_count_attr.attr,
+ &zcache_curr_obj_count_max_attr.attr,
+ &zcache_curr_objnode_count_attr.attr,
+ &zcache_curr_objnode_count_max_attr.attr,
+ &zcache_flush_total_attr.attr,
+ &zcache_flobj_total_attr.attr,
+ &zcache_flush_found_attr.attr,
+ &zcache_flobj_found_attr.attr,
+ &zcache_failed_eph_puts_attr.attr,
+ &zcache_nonactive_puts_attr.attr,
+ &zcache_failed_pers_puts_attr.attr,
+ &zcache_policy_percent_exceeded_attr.attr,
+ &zcache_compress_poor_attr.attr,
+ &zcache_mean_compress_poor_attr.attr,
+ &zcache_zbud_curr_raw_pages_attr.attr,
+ &zcache_zbud_curr_zpages_attr.attr,
+ &zcache_zbud_curr_zbytes_attr.attr,
+ &zcache_zbud_cumul_zpages_attr.attr,
+ &zcache_zbud_cumul_zbytes_attr.attr,
+ &zcache_zbud_buddied_count_attr.attr,
+ &zcache_evicted_raw_pages_attr.attr,
+ &zcache_evicted_unbuddied_pages_attr.attr,
+ &zcache_evicted_buddied_pages_attr.attr,
+ &zcache_failed_get_free_pages_attr.attr,
+ &zcache_failed_alloc_attr.attr,
+ &zcache_put_to_flush_attr.attr,
+ &zcache_zbud_unbuddied_list_counts_attr.attr,
+ &zcache_zbud_cumul_chunk_counts_attr.attr,
+ &zcache_zv_curr_dist_counts_attr.attr,
+ &zcache_zv_cumul_dist_counts_attr.attr,
+ &zcache_zv_max_zsize_attr.attr,
+ &zcache_zv_max_mean_zsize_attr.attr,
+ &zcache_zv_page_count_policy_percent_attr.attr,
+ NULL,
+};
+
+static struct attribute_group zcache_attr_group = {
+ .attrs = zcache_attrs,
+ .name = "zcache",
+};
+
+#define RAMSTER_SYSFS_RO(_name) \
+ static ssize_t ramster_##_name##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, char *buf) \
+ { \
+ return sprintf(buf, "%lu\n", ramster_##_name); \
+ } \
+ static struct kobj_attribute ramster_##_name##_attr = { \
+ .attr = { .name = __stringify(_name), .mode = 0444 }, \
+ .show = ramster_##_name##_show, \
+ }
+
+#define RAMSTER_SYSFS_RW(_name) \
+ static ssize_t ramster_##_name##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, char *buf) \
+ { \
+ return sprintf(buf, "%lu\n", ramster_##_name); \
+ } \
+ static ssize_t ramster_##_name##_store(struct kobject *kobj, \
+ struct kobj_attribute *attr, const char *buf, size_t count) \
+ { \
+ int err; \
+ unsigned long enable; \
+ err = kstrtoul(buf, 10, &enable); \
+ if (err) \
+ return -EINVAL; \
+ ramster_##_name = enable; \
+ return count; \
+ } \
+ static struct kobj_attribute ramster_##_name##_attr = { \
+ .attr = { .name = __stringify(_name), .mode = 0644 }, \
+ .show = ramster_##_name##_show, \
+ .store = ramster_##_name##_store, \
+ }
+
+#define RAMSTER_SYSFS_RO_ATOMIC(_name) \
+ static ssize_t ramster_##_name##_show(struct kobject *kobj, \
+ struct kobj_attribute *attr, char *buf) \
+ { \
+ return sprintf(buf, "%d\n", atomic_read(&ramster_##_name)); \
+ } \
+ static struct kobj_attribute ramster_##_name##_attr = { \
+ .attr = { .name = __stringify(_name), .mode = 0444 }, \
+ .show = ramster_##_name##_show, \
+ }
+
+RAMSTER_SYSFS_RO(interface_revision);
+RAMSTER_SYSFS_RO_ATOMIC(remote_pers_pages);
+RAMSTER_SYSFS_RW(pers_remotify_enable);
+RAMSTER_SYSFS_RW(eph_remotify_enable);
+RAMSTER_SYSFS_RO(eph_pages_remoted);
+RAMSTER_SYSFS_RO(eph_pages_remote_failed);
+RAMSTER_SYSFS_RO(pers_pages_remoted);
+RAMSTER_SYSFS_RO(pers_pages_remote_failed);
+RAMSTER_SYSFS_RO(pers_pages_remote_nomem);
+RAMSTER_SYSFS_RO(remote_pages_flushed);
+RAMSTER_SYSFS_RO(remote_page_flushes_failed);
+RAMSTER_SYSFS_RO(remote_objects_flushed);
+RAMSTER_SYSFS_RO(remote_object_flushes_failed);
+RAMSTER_SYSFS_RO(remote_eph_pages_succ_get);
+RAMSTER_SYSFS_RO(remote_eph_pages_unsucc_get);
+RAMSTER_SYSFS_RO(remote_pers_pages_succ_get);
+RAMSTER_SYSFS_RO(remote_pers_pages_unsucc_get);
+RAMSTER_SYSFS_RO_ATOMIC(foreign_eph_pampd_count);
+RAMSTER_SYSFS_RO(foreign_eph_pampd_count_max);
+RAMSTER_SYSFS_RO_ATOMIC(foreign_pers_pampd_count);
+RAMSTER_SYSFS_RO(foreign_pers_pampd_count_max);
+RAMSTER_SYSFS_RO_ATOMIC(curr_flnode_count);
+RAMSTER_SYSFS_RO(curr_flnode_count_max);
+
+#define MANUAL_NODES 8
+static bool ramster_nodes_manual_up[MANUAL_NODES];
+static ssize_t ramster_manual_node_up_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *buf)
+{
+ int i;
+ char *p = buf;
+ for (i = 0; i < MANUAL_NODES; i++)
+ if (ramster_nodes_manual_up[i])
+ p += sprintf(p, "%d ", i);
+ p += sprintf(p, "\n");
+ return p - buf;
+}
+
+static ssize_t ramster_manual_node_up_store(struct kobject *kobj,
+ struct kobj_attribute *attr, const char *buf, size_t count)
+{
+ int err;
+ unsigned long node_num;
+
+ err = kstrtoul(buf, 10, &node_num);
+ if (err) {
+ pr_err("ramster: bad strtoul?\n");
+ return -EINVAL;
+ }
+ if (node_num >= MANUAL_NODES) {
+ pr_err("ramster: bad node_num=%lu?\n", node_num);
+ return -EINVAL;
+ }
+ if (ramster_nodes_manual_up[node_num]) {
+ pr_err("ramster: node %d already up, ignoring\n",
+ (int)node_num);
+ } else {
+ ramster_nodes_manual_up[node_num] = true;
+ r2net_hb_node_up_manual((int)node_num);
+ }
+ return count;
+}
+
+static struct kobj_attribute ramster_manual_node_up_attr = {
+ .attr = { .name = "manual_node_up", .mode = 0644 },
+ .show = ramster_manual_node_up_show,
+ .store = ramster_manual_node_up_store,
+};
+
+static ssize_t ramster_remote_target_nodenum_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *buf)
+{
+ if (ramster_remote_target_nodenum == -1UL)
+ return sprintf(buf, "unset\n");
+ else
+ return sprintf(buf, "%d\n", ramster_remote_target_nodenum);
+}
+
+static ssize_t ramster_remote_target_nodenum_store(struct kobject *kobj,
+ struct kobj_attribute *attr, const char *buf, size_t count)
+{
+ int err;
+ unsigned long node_num;
+
+ err = kstrtoul(buf, 10, &node_num);
+ if (err) {
+ pr_err("ramster: bad strtoul?\n");
+ return -EINVAL;
+ } else if (node_num == -1UL) {
+ pr_err("ramster: disabling all remotification, "
+ "data may still reside on remote nodes however\n");
+ return -EINVAL;
+ } else if (node_num >= MANUAL_NODES) {
+ pr_err("ramster: bad node_num=%lu?\n", node_num);
+ return -EINVAL;
+ } else if (!ramster_nodes_manual_up[node_num]) {
+ pr_err("ramster: node %d not up, ignoring setting "
+ "of remotification target\n", (int)node_num);
+ } else if (r2net_remote_target_node_set((int)node_num) >= 0) {
+ pr_info("ramster: node %d set as remotification target\n",
+ (int)node_num);
+ ramster_remote_target_nodenum = (int)node_num;
+ } else {
+ pr_err("ramster: bad num to node node_num=%d?\n",
+ (int)node_num);
+ return -EINVAL;
+ }
+ return count;
+}
+
+static struct kobj_attribute ramster_remote_target_nodenum_attr = {
+ .attr = { .name = "remote_target_nodenum", .mode = 0644 },
+ .show = ramster_remote_target_nodenum_show,
+ .store = ramster_remote_target_nodenum_store,
+};
+
+
+static struct attribute *ramster_attrs[] = {
+ &ramster_interface_revision_attr.attr,
+ &ramster_pers_remotify_enable_attr.attr,
+ &ramster_eph_remotify_enable_attr.attr,
+ &ramster_remote_pers_pages_attr.attr,
+ &ramster_eph_pages_remoted_attr.attr,
+ &ramster_eph_pages_remote_failed_attr.attr,
+ &ramster_pers_pages_remoted_attr.attr,
+ &ramster_pers_pages_remote_failed_attr.attr,
+ &ramster_pers_pages_remote_nomem_attr.attr,
+ &ramster_remote_pages_flushed_attr.attr,
+ &ramster_remote_page_flushes_failed_attr.attr,
+ &ramster_remote_objects_flushed_attr.attr,
+ &ramster_remote_object_flushes_failed_attr.attr,
+ &ramster_remote_eph_pages_succ_get_attr.attr,
+ &ramster_remote_eph_pages_unsucc_get_attr.attr,
+ &ramster_remote_pers_pages_succ_get_attr.attr,
+ &ramster_remote_pers_pages_unsucc_get_attr.attr,
+ &ramster_foreign_eph_pampd_count_attr.attr,
+ &ramster_foreign_eph_pampd_count_max_attr.attr,
+ &ramster_foreign_pers_pampd_count_attr.attr,
+ &ramster_foreign_pers_pampd_count_max_attr.attr,
+ &ramster_curr_flnode_count_attr.attr,
+ &ramster_curr_flnode_count_max_attr.attr,
+ &ramster_manual_node_up_attr.attr,
+ &ramster_remote_target_nodenum_attr.attr,
+ NULL,
+};
+
+static struct attribute_group ramster_attr_group = {
+ .attrs = ramster_attrs,
+ .name = "ramster",
+};
+
+#endif /* CONFIG_SYSFS */
+/*
+ * When zcache is disabled ("frozen"), pools can be created and destroyed,
+ * but all puts (and thus all other operations that require memory allocation)
+ * must fail. If zcache is unfrozen, accepts puts, then frozen again,
+ * data consistency requires all puts while frozen to be converted into
+ * flushes.
+ */
+static bool zcache_freeze;
+
+/*
+ * zcache shrinker interface (only useful for ephemeral pages, so zbud only)
+ */
+static int shrink_zcache_memory(struct shrinker *shrink,
+ struct shrink_control *sc)
+{
+ int ret = -1;
+ int nr = sc->nr_to_scan;
+ gfp_t gfp_mask = sc->gfp_mask;
+
+ if (nr >= 0) {
+ if (!(gfp_mask & __GFP_FS))
+ /* does this case really need to be skipped? */
+ goto out;
+ zbud_evict_pages(nr);
+ }
+ ret = (int)atomic_read(&zcache_zbud_curr_raw_pages);
+out:
+ return ret;
+}
+
+static struct shrinker zcache_shrinker = {
+ .shrink = shrink_zcache_memory,
+ .seeks = DEFAULT_SEEKS,
+};
+
+/*
+ * zcache shims between cleancache/frontswap ops and tmem
+ */
+
+int zcache_put(int cli_id, int pool_id, struct tmem_oid *oidp,
+ uint32_t index, char *data, size_t size,
+ bool raw, int ephemeral)
+{
+ struct tmem_pool *pool;
+ int ret = -1;
+
+ BUG_ON(!irqs_disabled());
+ pool = zcache_get_pool_by_id(cli_id, pool_id);
+ if (unlikely(pool == NULL))
+ goto out;
+ if (!zcache_freeze && zcache_do_preload(pool) == 0) {
+ /* preload does preempt_disable on success */
+ ret = tmem_put(pool, oidp, index, data, size, raw, ephemeral);
+ if (ret < 0) {
+ if (is_ephemeral(pool))
+ zcache_failed_eph_puts++;
+ else
+ zcache_failed_pers_puts++;
+ }
+ zcache_put_pool(pool);
+ preempt_enable_no_resched();
+ } else {
+ zcache_put_to_flush++;
+ if (atomic_read(&pool->obj_count) > 0)
+ /* the put fails whether the flush succeeds or not */
+ (void)tmem_flush_page(pool, oidp, index);
+ zcache_put_pool(pool);
+ }
+out:
+ return ret;
+}
+
+int zcache_get(int cli_id, int pool_id, struct tmem_oid *oidp,
+ uint32_t index, char *data, size_t *sizep,
+ bool raw, int get_and_free)
+{
+ struct tmem_pool *pool;
+ int ret = -1;
+ bool eph;
+
+ if (!raw) {
+ BUG_ON(irqs_disabled());
+ BUG_ON(in_softirq());
+ }
+ pool = zcache_get_pool_by_id(cli_id, pool_id);
+ eph = is_ephemeral(pool);
+ if (likely(pool != NULL)) {
+ if (atomic_read(&pool->obj_count) > 0)
+ ret = tmem_get(pool, oidp, index, data, sizep,
+ raw, get_and_free);
+ zcache_put_pool(pool);
+ }
+ WARN_ONCE((!eph && (ret != 0)), "zcache_get fails on persistent pool, "
+ "bad things are very likely to happen soon\n");
+#ifdef RAMSTER_TESTING
+ if (ret != 0 && ret != -1 && !(ret == -EINVAL && is_ephemeral(pool)))
+ pr_err("TESTING zcache_get tmem_get returns ret=%d\n", ret);
+#endif
+ if (ret == -EAGAIN)
+ BUG(); /* FIXME... don't need this anymore??? let's ensure */
+ return ret;
+}
+
+int zcache_flush(int cli_id, int pool_id,
+ struct tmem_oid *oidp, uint32_t index)
+{
+ struct tmem_pool *pool;
+ int ret = -1;
+ unsigned long flags;
+
+ local_irq_save(flags);
+ zcache_flush_total++;
+ pool = zcache_get_pool_by_id(cli_id, pool_id);
+ ramster_do_preload_flnode_only(pool);
+ if (likely(pool != NULL)) {
+ if (atomic_read(&pool->obj_count) > 0)
+ ret = tmem_flush_page(pool, oidp, index);
+ zcache_put_pool(pool);
+ }
+ if (ret >= 0)
+ zcache_flush_found++;
+ local_irq_restore(flags);
+ return ret;
+}
+
+int zcache_flush_object(int cli_id, int pool_id, struct tmem_oid *oidp)
+{
+ struct tmem_pool *pool;
+ int ret = -1;
+ unsigned long flags;
+
+ local_irq_save(flags);
+ zcache_flobj_total++;
+ pool = zcache_get_pool_by_id(cli_id, pool_id);
+ ramster_do_preload_flnode_only(pool);
+ if (likely(pool != NULL)) {
+ if (atomic_read(&pool->obj_count) > 0)
+ ret = tmem_flush_object(pool, oidp);
+ zcache_put_pool(pool);
+ }
+ if (ret >= 0)
+ zcache_flobj_found++;
+ local_irq_restore(flags);
+ return ret;
+}
+
+int zcache_client_destroy_pool(int cli_id, int pool_id)
+{
+ struct tmem_pool *pool = NULL;
+ struct zcache_client *cli = NULL;
+ int ret = -1;
+
+ if (pool_id < 0)
+ goto out;
+ if (cli_id == LOCAL_CLIENT)
+ cli = &zcache_host;
+ else if ((unsigned int)cli_id < MAX_CLIENTS)
+ cli = &zcache_clients[cli_id];
+ if (cli == NULL)
+ goto out;
+ atomic_inc(&cli->refcount);
+ pool = cli->tmem_pools[pool_id];
+ if (pool == NULL)
+ goto out;
+ cli->tmem_pools[pool_id] = NULL;
+ /* wait for pool activity on other cpus to quiesce */
+ while (atomic_read(&pool->refcount) != 0)
+ ;
+ atomic_dec(&cli->refcount);
+ local_bh_disable();
+ ret = tmem_destroy_pool(pool);
+ local_bh_enable();
+ kfree(pool);
+ pr_info("ramster: destroyed pool id=%d cli_id=%d\n", pool_id, cli_id);
+out:
+ return ret;
+}
+
+static int zcache_destroy_pool(int pool_id)
+{
+ return zcache_client_destroy_pool(LOCAL_CLIENT, pool_id);
+}
+
+int zcache_new_pool(uint16_t cli_id, uint32_t flags)
+{
+ int poolid = -1;
+ struct tmem_pool *pool;
+ struct zcache_client *cli = NULL;
+
+ if (cli_id == LOCAL_CLIENT)
+ cli = &zcache_host;
+ else if ((unsigned int)cli_id < MAX_CLIENTS)
+ cli = &zcache_clients[cli_id];
+ if (cli == NULL)
+ goto out;
+ atomic_inc(&cli->refcount);
+ pool = kmalloc(sizeof(struct tmem_pool), GFP_ATOMIC);
+ if (pool == NULL) {
+ pr_info("ramster: pool creation failed: out of memory\n");
+ goto out;
+ }
+
+ for (poolid = 0; poolid < MAX_POOLS_PER_CLIENT; poolid++)
+ if (cli->tmem_pools[poolid] == NULL)
+ break;
+ if (poolid >= MAX_POOLS_PER_CLIENT) {
+ pr_info("ramster: pool creation failed: max exceeded\n");
+ kfree(pool);
+ poolid = -1;
+ goto out;
+ }
+ atomic_set(&pool->refcount, 0);
+ pool->client = cli;
+ pool->pool_id = poolid;
+ tmem_new_pool(pool, flags);
+ cli->tmem_pools[poolid] = pool;
+ if (cli_id == LOCAL_CLIENT)
+ pr_info("ramster: created %s tmem pool, id=%d, local client\n",
+ flags & TMEM_POOL_PERSIST ? "persistent" : "ephemeral",
+ poolid);
+ else
+ pr_info("ramster: created %s tmem pool, id=%d, client=%d\n",
+ flags & TMEM_POOL_PERSIST ? "persistent" : "ephemeral",
+ poolid, cli_id);
+out:
+ if (cli != NULL)
+ atomic_dec(&cli->refcount);
+ return poolid;
+}
+
+static int zcache_local_new_pool(uint32_t flags)
+{
+ return zcache_new_pool(LOCAL_CLIENT, flags);
+}
+
+int zcache_autocreate_pool(int cli_id, int pool_id, bool ephemeral)
+{
+ struct tmem_pool *pool;
+ struct zcache_client *cli = NULL;
+ uint32_t flags = ephemeral ? 0 : TMEM_POOL_PERSIST;
+ int ret = -1;
+
+ if (cli_id == LOCAL_CLIENT)
+ goto out;
+ if (pool_id >= MAX_POOLS_PER_CLIENT)
+ goto out;
+ else if ((unsigned int)cli_id < MAX_CLIENTS)
+ cli = &zcache_clients[cli_id];
+ if ((ephemeral && !use_cleancache) || (!ephemeral && !use_frontswap))
+ BUG(); /* FIXME, handle more gracefully later */
+ if (!cli->allocated) {
+ if (zcache_new_client(cli_id))
+ BUG(); /* FIXME, handle more gracefully later */
+ cli = &zcache_clients[cli_id];
+ }
+ atomic_inc(&cli->refcount);
+ pool = cli->tmem_pools[pool_id];
+ if (pool != NULL) {
+ if (pool->persistent && ephemeral) {
+ pr_err("zcache_autocreate_pool: type mismatch\n");
+ goto out;
+ }
+ ret = 0;
+ goto out;
+ }
+ pool = kmalloc(sizeof(struct tmem_pool), GFP_KERNEL);
+ if (pool == NULL) {
+ pr_info("ramster: pool creation failed: out of memory\n");
+ goto out;
+ }
+ atomic_set(&pool->refcount, 0);
+ pool->client = cli;
+ pool->pool_id = pool_id;
+ tmem_new_pool(pool, flags);
+ cli->tmem_pools[pool_id] = pool;
+ pr_info("ramster: AUTOcreated %s tmem poolid=%d, for remote client=%d\n",
+ flags & TMEM_POOL_PERSIST ? "persistent" : "ephemeral",
+ pool_id, cli_id);
+ ret = 0;
+out:
+ if (cli == NULL)
+ BUG(); /* FIXME, handle more gracefully later */
+ /* pr_err("zcache_autocreate_pool: failed\n"); */
+ if (cli != NULL)
+ atomic_dec(&cli->refcount);
+ return ret;
+}
+
+/**********
+ * Two kernel functionalities currently can be layered on top of tmem.
+ * These are "cleancache" which is used as a second-chance cache for clean
+ * page cache pages; and "frontswap" which is used for swap pages
+ * to avoid writes to disk. A generic "shim" is provided here for each
+ * to translate in-kernel semantics to zcache semantics.
+ */
+
+#ifdef CONFIG_CLEANCACHE
+static void zcache_cleancache_put_page(int pool_id,
+ struct cleancache_filekey key,
+ pgoff_t index, struct page *page)
+{
+ u32 ind = (u32) index;
+ struct tmem_oid oid = *(struct tmem_oid *)&key;
+
+#ifdef __PG_WAS_ACTIVE
+ if (!PageWasActive(page)) {
+ zcache_nonactive_puts++;
+ return;
+ }
+#endif
+ if (likely(ind == index)) {
+ char *kva = page_address(page);
+
+ (void)zcache_put(LOCAL_CLIENT, pool_id, &oid, index,
+ kva, PAGE_SIZE, 0, 1);
+ }
+}
+
+static int zcache_cleancache_get_page(int pool_id,
+ struct cleancache_filekey key,
+ pgoff_t index, struct page *page)
+{
+ u32 ind = (u32) index;
+ struct tmem_oid oid = *(struct tmem_oid *)&key;
+ int ret = -1;
+
+ preempt_disable();
+ if (likely(ind == index)) {
+ char *kva = page_address(page);
+ size_t size = PAGE_SIZE;
+
+ ret = zcache_get(LOCAL_CLIENT, pool_id, &oid, index,
+ kva, &size, 0, 0);
+#ifdef __PG_WAS_ACTIVE
+ if (ret == 0)
+ SetPageWasActive(page);
+#endif
+ }
+ preempt_enable();
+ return ret;
+}
+
+static void zcache_cleancache_flush_page(int pool_id,
+ struct cleancache_filekey key,
+ pgoff_t index)
+{
+ u32 ind = (u32) index;
+ struct tmem_oid oid = *(struct tmem_oid *)&key;
+
+ if (likely(ind == index))
+ (void)zcache_flush(LOCAL_CLIENT, pool_id, &oid, ind);
+}
+
+static void zcache_cleancache_flush_inode(int pool_id,
+ struct cleancache_filekey key)
+{
+ struct tmem_oid oid = *(struct tmem_oid *)&key;
+
+ (void)zcache_flush_object(LOCAL_CLIENT, pool_id, &oid);
+}
+
+static void zcache_cleancache_flush_fs(int pool_id)
+{
+ if (pool_id >= 0)
+ (void)zcache_destroy_pool(pool_id);
+}
+
+static int zcache_cleancache_init_fs(size_t pagesize)
+{
+ BUG_ON(sizeof(struct cleancache_filekey) !=
+ sizeof(struct tmem_oid));
+ BUG_ON(pagesize != PAGE_SIZE);
+ return zcache_local_new_pool(0);
+}
+
+static int zcache_cleancache_init_shared_fs(char *uuid, size_t pagesize)
+{
+ /* shared pools are unsupported and map to private */
+ BUG_ON(sizeof(struct cleancache_filekey) !=
+ sizeof(struct tmem_oid));
+ BUG_ON(pagesize != PAGE_SIZE);
+ return zcache_local_new_pool(0);
+}
+
+static struct cleancache_ops zcache_cleancache_ops = {
+ .put_page = zcache_cleancache_put_page,
+ .get_page = zcache_cleancache_get_page,
+ .invalidate_page = zcache_cleancache_flush_page,
+ .invalidate_inode = zcache_cleancache_flush_inode,
+ .invalidate_fs = zcache_cleancache_flush_fs,
+ .init_shared_fs = zcache_cleancache_init_shared_fs,
+ .init_fs = zcache_cleancache_init_fs
+};
+
+struct cleancache_ops zcache_cleancache_register_ops(void)
+{
+ struct cleancache_ops old_ops =
+ cleancache_register_ops(&zcache_cleancache_ops);
+
+ return old_ops;
+}
+#endif
+
+#ifdef CONFIG_FRONTSWAP
+/* a single tmem poolid is used for all frontswap "types" (swapfiles) */
+static int zcache_frontswap_poolid = -1;
+
+/*
+ * Swizzling increases objects per swaptype, increasing tmem concurrency
+ * for heavy swaploads. Later, larger nr_cpus -> larger SWIZ_BITS
+ */
+#define SWIZ_BITS 8
+#define SWIZ_MASK ((1 << SWIZ_BITS) - 1)
+#define _oswiz(_type, _ind) ((_type << SWIZ_BITS) | (_ind & SWIZ_MASK))
+#define iswiz(_ind) (_ind >> SWIZ_BITS)
+
+static inline struct tmem_oid oswiz(unsigned type, u32 ind)
+{
+ struct tmem_oid oid = { .oid = { 0 } };
+ oid.oid[0] = _oswiz(type, ind);
+ return oid;
+}
+
+static int zcache_frontswap_put_page(unsigned type, pgoff_t offset,
+ struct page *page)
+{
+ u64 ind64 = (u64)offset;
+ u32 ind = (u32)offset;
+ struct tmem_oid oid = oswiz(type, ind);
+ int ret = -1;
+ unsigned long flags;
+ char *kva;
+
+ BUG_ON(!PageLocked(page));
+ if (likely(ind64 == ind)) {
+ local_irq_save(flags);
+ kva = page_address(page);
+ ret = zcache_put(LOCAL_CLIENT, zcache_frontswap_poolid,
+ &oid, iswiz(ind), kva, PAGE_SIZE, 0, 0);
+ local_irq_restore(flags);
+ }
+ return ret;
+}
+
+/* returns 0 if the page was successfully gotten from frontswap, -1 if
+ * was not present (should never happen!) */
+static int zcache_frontswap_get_page(unsigned type, pgoff_t offset,
+ struct page *page)
+{
+ u64 ind64 = (u64)offset;
+ u32 ind = (u32)offset;
+ struct tmem_oid oid = oswiz(type, ind);
+ int ret = -1;
+
+ preempt_disable(); /* FIXME, remove this? */
+ BUG_ON(!PageLocked(page));
+ if (likely(ind64 == ind)) {
+ char *kva = page_address(page);
+ size_t size = PAGE_SIZE;
+
+ ret = zcache_get(LOCAL_CLIENT, zcache_frontswap_poolid,
+ &oid, iswiz(ind), kva, &size, 0, -1);
+ }
+ preempt_enable(); /* FIXME, remove this? */
+ return ret;
+}
+
+/* flush a single page from frontswap */
+static void zcache_frontswap_flush_page(unsigned type, pgoff_t offset)
+{
+ u64 ind64 = (u64)offset;
+ u32 ind = (u32)offset;
+ struct tmem_oid oid = oswiz(type, ind);
+
+ if (likely(ind64 == ind))
+ (void)zcache_flush(LOCAL_CLIENT, zcache_frontswap_poolid,
+ &oid, iswiz(ind));
+}
+
+/* flush all pages from the passed swaptype */
+static void zcache_frontswap_flush_area(unsigned type)
+{
+ struct tmem_oid oid;
+ int ind;
+
+ for (ind = SWIZ_MASK; ind >= 0; ind--) {
+ oid = oswiz(type, ind);
+ (void)zcache_flush_object(LOCAL_CLIENT,
+ zcache_frontswap_poolid, &oid);
+ }
+}
+
+static void zcache_frontswap_init(unsigned ignored)
+{
+ /* a single tmem poolid is used for all frontswap "types" (swapfiles) */
+ if (zcache_frontswap_poolid < 0)
+ zcache_frontswap_poolid =
+ zcache_local_new_pool(TMEM_POOL_PERSIST);
+}
+
+static struct frontswap_ops zcache_frontswap_ops = {
+ .put_page = zcache_frontswap_put_page,
+ .get_page = zcache_frontswap_get_page,
+ .invalidate_page = zcache_frontswap_flush_page,
+ .invalidate_area = zcache_frontswap_flush_area,
+ .init = zcache_frontswap_init
+};
+
+struct frontswap_ops zcache_frontswap_register_ops(void)
+{
+ struct frontswap_ops old_ops =
+ frontswap_register_ops(&zcache_frontswap_ops);
+
+ return old_ops;
+}
+#endif
+
+/*
+ * frontswap selfshrinking
+ */
+
+#ifdef CONFIG_FRONTSWAP
+/* In HZ, controls frequency of worker invocation. */
+static unsigned int selfshrink_interval __read_mostly = 5;
+
+static void selfshrink_process(struct work_struct *work);
+static DECLARE_DELAYED_WORK(selfshrink_worker, selfshrink_process);
+
+/* Enable/disable with sysfs. */
+static bool frontswap_selfshrinking __read_mostly;
+
+/* Enable/disable with kernel boot option. */
+static bool use_frontswap_selfshrink __initdata = true;
+
+/*
+ * The default values for the following parameters were deemed reasonable
+ * by experimentation, may be workload-dependent, and can all be
+ * adjusted via sysfs.
+ */
+
+/* Control rate for frontswap shrinking. Higher hysteresis is slower. */
+static unsigned int frontswap_hysteresis __read_mostly = 20;
+
+/*
+ * Number of selfshrink worker invocations to wait before observing that
+ * frontswap selfshrinking should commence. Note that selfshrinking does
+ * not use a separate worker thread.
+ */
+static unsigned int frontswap_inertia __read_mostly = 3;
+
+/* Countdown to next invocation of frontswap_shrink() */
+static unsigned long frontswap_inertia_counter;
+
+/*
+ * Invoked by the selfshrink worker thread, uses current number of pages
+ * in frontswap (frontswap_curr_pages()), previous status, and control
+ * values (hysteresis and inertia) to determine if frontswap should be
+ * shrunk and what the new frontswap size should be. Note that
+ * frontswap_shrink is essentially a partial swapoff that immediately
+ * transfers pages from the "swap device" (frontswap) back into kernel
+ * RAM; despite the name, frontswap "shrinking" is very different from
+ * the "shrinker" interface used by the kernel MM subsystem to reclaim
+ * memory.
+ */
+static void frontswap_selfshrink(void)
+{
+ static unsigned long cur_frontswap_pages;
+ static unsigned long last_frontswap_pages;
+ static unsigned long tgt_frontswap_pages;
+
+ last_frontswap_pages = cur_frontswap_pages;
+ cur_frontswap_pages = frontswap_curr_pages();
+ if (!cur_frontswap_pages ||
+ (cur_frontswap_pages > last_frontswap_pages)) {
+ frontswap_inertia_counter = frontswap_inertia;
+ return;
+ }
+ if (frontswap_inertia_counter && --frontswap_inertia_counter)
+ return;
+ if (cur_frontswap_pages <= frontswap_hysteresis)
+ tgt_frontswap_pages = 0;
+ else
+ tgt_frontswap_pages = cur_frontswap_pages -
+ (cur_frontswap_pages / frontswap_hysteresis);
+ frontswap_shrink(tgt_frontswap_pages);
+}
+
+static int __init ramster_nofrontswap_selfshrink_setup(char *s)
+{
+ use_frontswap_selfshrink = false;
+ return 1;
+}
+
+__setup("noselfshrink", ramster_nofrontswap_selfshrink_setup);
+
+static void selfshrink_process(struct work_struct *work)
+{
+ if (frontswap_selfshrinking && frontswap_enabled) {
+ frontswap_selfshrink();
+ schedule_delayed_work(&selfshrink_worker,
+ selfshrink_interval * HZ);
+ }
+}
+
+static int ramster_enabled;
+
+static int __init ramster_selfshrink_init(void)
+{
+ frontswap_selfshrinking = ramster_enabled && use_frontswap_selfshrink;
+ if (frontswap_selfshrinking)
+ pr_info("ramster: Initializing frontswap "
+ "selfshrinking driver.\n");
+ else
+ return -ENODEV;
+
+ schedule_delayed_work(&selfshrink_worker, selfshrink_interval * HZ);
+
+ return 0;
+}
+
+subsys_initcall(ramster_selfshrink_init);
+#endif
+
+/*
+ * zcache initialization
+ * NOTE FOR NOW ramster MUST BE PROVIDED AS A KERNEL BOOT PARAMETER OR
+ * NOTHING HAPPENS!
+ */
+
+static int ramster_enabled;
+
+static int __init enable_ramster(char *s)
+{
+ ramster_enabled = 1;
+ return 1;
+}
+__setup("ramster", enable_ramster);
+
+/* allow independent dynamic disabling of cleancache and frontswap */
+
+static int use_cleancache = 1;
+
+static int __init no_cleancache(char *s)
+{
+ pr_info("INIT no_cleancache called\n");
+ use_cleancache = 0;
+ return 1;
+}
+
+/*
+ * FIXME: need to guarantee this gets checked before zcache_init is called
+ * What is the correct way to achieve this?
+ */
+early_param("nocleancache", no_cleancache);
+
+static int use_frontswap = 1;
+
+static int __init no_frontswap(char *s)
+{
+ pr_info("INIT no_frontswap called\n");
+ use_frontswap = 0;
+ return 1;
+}
+
+__setup("nofrontswap", no_frontswap);
+
+static int __init zcache_init(void)
+{
+ int ret = 0;
+
+#ifdef CONFIG_SYSFS
+ ret = sysfs_create_group(mm_kobj, &zcache_attr_group);
+ ret = sysfs_create_group(mm_kobj, &ramster_attr_group);
+ if (ret) {
+ pr_err("ramster: can't create sysfs\n");
+ goto out;
+ }
+#endif /* CONFIG_SYSFS */
+#if defined(CONFIG_CLEANCACHE) || defined(CONFIG_FRONTSWAP)
+ if (ramster_enabled) {
+ unsigned int cpu;
+
+ (void)r2net_register_handlers();
+ tmem_register_hostops(&zcache_hostops);
+ tmem_register_pamops(&zcache_pamops);
+ ret = register_cpu_notifier(&zcache_cpu_notifier_block);
+ if (ret) {
+ pr_err("ramster: can't register cpu notifier\n");
+ goto out;
+ }
+ for_each_online_cpu(cpu) {
+ void *pcpu = (void *)(long)cpu;
+ zcache_cpu_notifier(&zcache_cpu_notifier_block,
+ CPU_UP_PREPARE, pcpu);
+ }
+ }
+ zcache_objnode_cache = kmem_cache_create("zcache_objnode",
+ sizeof(struct tmem_objnode), 0, 0, NULL);
+ zcache_obj_cache = kmem_cache_create("zcache_obj",
+ sizeof(struct tmem_obj), 0, 0, NULL);
+ ramster_flnode_cache = kmem_cache_create("ramster_flnode",
+ sizeof(struct flushlist_node), 0, 0, NULL);
+#endif
+#ifdef CONFIG_CLEANCACHE
+ pr_info("INIT ramster_enabled=%d use_cleancache=%d\n",
+ ramster_enabled, use_cleancache);
+ if (ramster_enabled && use_cleancache) {
+ struct cleancache_ops old_ops;
+
+ zbud_init();
+ register_shrinker(&zcache_shrinker);
+ old_ops = zcache_cleancache_register_ops();
+ pr_info("ramster: cleancache enabled using kernel "
+ "transcendent memory and compression buddies\n");
+ if (old_ops.init_fs != NULL)
+ pr_warning("ramster: cleancache_ops overridden");
+ }
+#endif
+#ifdef CONFIG_FRONTSWAP
+ pr_info("INIT ramster_enabled=%d use_frontswap=%d\n",
+ ramster_enabled, use_frontswap);
+ if (ramster_enabled && use_frontswap) {
+ struct frontswap_ops old_ops;
+
+ zcache_new_client(LOCAL_CLIENT);
+ old_ops = zcache_frontswap_register_ops();
+ pr_info("ramster: frontswap enabled using kernel "
+ "transcendent memory and xvmalloc\n");
+ if (old_ops.init != NULL)
+ pr_warning("ramster: frontswap_ops overridden");
+ }
+ if (ramster_enabled && (use_frontswap || use_cleancache))
+ ramster_remotify_init();
+#endif
+out:
+ return ret;
+}
+
+module_init(zcache_init)
*
* Zcache provides an in-kernel "host implementation" for transcendent memory
* and, thus indirectly, for cleancache and frontswap. Zcache includes two
- * page-accessible memory [1] interfaces, both utilizing lzo1x compression:
+ * page-accessible memory [1] interfaces, both utilizing the crypto compression
+ * API:
* 1) "compression buddies" ("zbud") is used for ephemeral pages
- * 2) xvmalloc is used for persistent pages.
+ * 2) zsmalloc is used for persistent pages.
* Xvmalloc (based on the TLSF allocator) has very low fragmentation
* so maximizes space efficiency, while zbud allows pairs (and potentially,
* in the future, more than a pair of) compressed pages to be closely linked
#include <linux/cpu.h>
#include <linux/highmem.h>
#include <linux/list.h>
-#include <linux/lzo.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/types.h>
#include <linux/atomic.h>
#include <linux/math64.h>
+#include <linux/crypto.h>
+#include <linux/string.h>
#include "tmem.h"
-#include "../zram/xvmalloc.h" /* if built in drivers/staging */
+#include "../zsmalloc/zsmalloc.h"
#if (!defined(CONFIG_CLEANCACHE) && !defined(CONFIG_FRONTSWAP))
#error "zcache is useless without CONFIG_CLEANCACHE or CONFIG_FRONTSWAP"
struct zcache_client {
struct tmem_pool *tmem_pools[MAX_POOLS_PER_CLIENT];
- struct xv_pool *xvpool;
+ struct zs_pool *zspool;
bool allocated;
atomic_t refcount;
};
return cli == &zcache_host;
}
+/* crypto API for zcache */
+#define ZCACHE_COMP_NAME_SZ CRYPTO_MAX_ALG_NAME
+static char zcache_comp_name[ZCACHE_COMP_NAME_SZ];
+static struct crypto_comp * __percpu *zcache_comp_pcpu_tfms;
+
+enum comp_op {
+ ZCACHE_COMPOP_COMPRESS,
+ ZCACHE_COMPOP_DECOMPRESS
+};
+
+static inline int zcache_comp_op(enum comp_op op,
+ const u8 *src, unsigned int slen,
+ u8 *dst, unsigned int *dlen)
+{
+ struct crypto_comp *tfm;
+ int ret;
+
+ BUG_ON(!zcache_comp_pcpu_tfms);
+ tfm = *per_cpu_ptr(zcache_comp_pcpu_tfms, get_cpu());
+ BUG_ON(!tfm);
+ switch (op) {
+ case ZCACHE_COMPOP_COMPRESS:
+ ret = crypto_comp_compress(tfm, src, slen, dst, dlen);
+ break;
+ case ZCACHE_COMPOP_DECOMPRESS:
+ ret = crypto_comp_decompress(tfm, src, slen, dst, dlen);
+ break;
+ }
+ put_cpu();
+ return ret;
+}
+
/**********
* Compression buddies ("zbud") provides for packing two (or, possibly
* in the future, more) compressed ephemeral pages into a single "raw"
struct zbud_page *zbpg =
container_of(zh, struct zbud_page, buddy[budnum]);
+ spin_lock(&zbud_budlists_spinlock);
spin_lock(&zbpg->lock);
if (list_empty(&zbpg->bud_list)) {
/* ignore zombie page... see zbud_evict_pages() */
spin_unlock(&zbpg->lock);
+ spin_unlock(&zbud_budlists_spinlock);
return;
}
size = zbud_free(zh);
zh_other = &zbpg->buddy[(budnum == 0) ? 1 : 0];
if (zh_other->size == 0) { /* was unbuddied: unlist and free */
chunks = zbud_size_to_chunks(size) ;
- spin_lock(&zbud_budlists_spinlock);
BUG_ON(list_empty(&zbud_unbuddied[chunks].list));
list_del_init(&zbpg->bud_list);
zbud_unbuddied[chunks].count--;
zbud_free_raw_page(zbpg);
} else { /* was buddied: move remaining buddy to unbuddied list */
chunks = zbud_size_to_chunks(zh_other->size) ;
- spin_lock(&zbud_budlists_spinlock);
list_del_init(&zbpg->bud_list);
zcache_zbud_buddied_count--;
list_add_tail(&zbpg->bud_list, &zbud_unbuddied[chunks].list);
{
struct zbud_page *zbpg;
unsigned budnum = zbud_budnum(zh);
- size_t out_len = PAGE_SIZE;
+ unsigned int out_len = PAGE_SIZE;
char *to_va, *from_va;
unsigned size;
int ret = 0;
}
ASSERT_SENTINEL(zh, ZBH);
BUG_ON(zh->size == 0 || zh->size > zbud_max_buddy_size());
- to_va = kmap_atomic(page, KM_USER0);
+ to_va = kmap_atomic(page);
size = zh->size;
from_va = zbud_data(zh, size);
- ret = lzo1x_decompress_safe(from_va, size, to_va, &out_len);
- BUG_ON(ret != LZO_E_OK);
+ ret = zcache_comp_op(ZCACHE_COMPOP_DECOMPRESS, from_va, size,
+ to_va, &out_len);
+ BUG_ON(ret);
BUG_ON(out_len != PAGE_SIZE);
- kunmap_atomic(to_va, KM_USER0);
+ kunmap_atomic(to_va);
out:
spin_unlock(&zbpg->lock);
return ret;
#endif
/**********
- * This "zv" PAM implementation combines the TLSF-based xvMalloc
- * with lzo1x compression to maximize the amount of data that can
+ * This "zv" PAM implementation combines the slab-based zsmalloc
+ * with the crypto compression API to maximize the amount of data that can
* be packed into a physical page.
*
* Zv represents a PAM page with the index and object (plus a "size" value
uint32_t pool_id;
struct tmem_oid oid;
uint32_t index;
+ size_t size;
DECL_SENTINEL
};
static atomic_t zv_curr_dist_counts[NCHUNKS];
static atomic_t zv_cumul_dist_counts[NCHUNKS];
-static struct zv_hdr *zv_create(struct xv_pool *xvpool, uint32_t pool_id,
+static struct zv_hdr *zv_create(struct zs_pool *pool, uint32_t pool_id,
struct tmem_oid *oid, uint32_t index,
void *cdata, unsigned clen)
{
- struct page *page;
- struct zv_hdr *zv = NULL;
- uint32_t offset;
- int alloc_size = clen + sizeof(struct zv_hdr);
- int chunks = (alloc_size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
- int ret;
+ struct zv_hdr *zv;
+ u32 size = clen + sizeof(struct zv_hdr);
+ int chunks = (size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
+ void *handle = NULL;
BUG_ON(!irqs_disabled());
BUG_ON(chunks >= NCHUNKS);
- ret = xv_malloc(xvpool, alloc_size,
- &page, &offset, ZCACHE_GFP_MASK);
- if (unlikely(ret))
+ handle = zs_malloc(pool, size);
+ if (!handle)
goto out;
atomic_inc(&zv_curr_dist_counts[chunks]);
atomic_inc(&zv_cumul_dist_counts[chunks]);
- zv = kmap_atomic(page) + offset;
+ zv = zs_map_object(pool, handle);
zv->index = index;
zv->oid = *oid;
zv->pool_id = pool_id;
+ zv->size = clen;
SET_SENTINEL(zv, ZVH);
memcpy((char *)zv + sizeof(struct zv_hdr), cdata, clen);
- kunmap_atomic(zv);
+ zs_unmap_object(pool, handle);
out:
- return zv;
+ return handle;
}
-static void zv_free(struct xv_pool *xvpool, struct zv_hdr *zv)
+static void zv_free(struct zs_pool *pool, void *handle)
{
unsigned long flags;
- struct page *page;
- uint32_t offset;
- uint16_t size = xv_get_object_size(zv);
- int chunks = (size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
+ struct zv_hdr *zv;
+ uint16_t size;
+ int chunks;
+ zv = zs_map_object(pool, handle);
ASSERT_SENTINEL(zv, ZVH);
+ size = zv->size + sizeof(struct zv_hdr);
+ INVERT_SENTINEL(zv, ZVH);
+ zs_unmap_object(pool, handle);
+
+ chunks = (size + (CHUNK_SIZE - 1)) >> CHUNK_SHIFT;
BUG_ON(chunks >= NCHUNKS);
atomic_dec(&zv_curr_dist_counts[chunks]);
- size -= sizeof(*zv);
- BUG_ON(size == 0);
- INVERT_SENTINEL(zv, ZVH);
- page = virt_to_page(zv);
- offset = (unsigned long)zv & ~PAGE_MASK;
+
local_irq_save(flags);
- xv_free(xvpool, page, offset);
+ zs_free(pool, handle);
local_irq_restore(flags);
}
-static void zv_decompress(struct page *page, struct zv_hdr *zv)
+static void zv_decompress(struct page *page, void *handle)
{
- size_t clen = PAGE_SIZE;
+ unsigned int clen = PAGE_SIZE;
char *to_va;
- unsigned size;
int ret;
+ struct zv_hdr *zv;
+ zv = zs_map_object(zcache_host.zspool, handle);
+ BUG_ON(zv->size == 0);
ASSERT_SENTINEL(zv, ZVH);
- to_va = kmap_atomic(page, KM_USER0);
- size = xv_get_object_size(zv) - sizeof(*zv);
- BUG_ON(size == 0);
+ to_va = kmap_atomic(page);
- ret = lzo1x_decompress_safe((char *)zv + sizeof(*zv),
- size, to_va, &clen);
+ ret = zcache_comp_op(ZCACHE_COMPOP_DECOMPRESS, (char *)zv + sizeof(*zv),
+ zv->size, to_va, &clen);
- kunmap_atomic(to_va, KM_USER0);
+ kunmap_atomic(to_va);
- BUG_ON(ret != LZO_E_OK);
+ zs_unmap_object(zcache_host.zspool, handle);
+ BUG_ON(ret);
BUG_ON(clen != PAGE_SIZE);
}
goto out;
cli->allocated = 1;
#ifdef CONFIG_FRONTSWAP
- cli->xvpool = xv_create_pool();
- if (cli->xvpool == NULL)
+ cli->zspool = zs_create_pool("zcache", ZCACHE_GFP_MASK);
+ if (cli->zspool == NULL)
goto out;
#endif
ret = 0;
static unsigned long zcache_curr_pers_pampd_count_max;
/* forward reference */
-static int zcache_compress(struct page *from, void **out_va, size_t *out_len);
+static int zcache_compress(struct page *from, void **out_va, unsigned *out_len);
static void *zcache_pampd_create(char *data, size_t size, bool raw, int eph,
struct tmem_pool *pool, struct tmem_oid *oid,
uint32_t index)
{
void *pampd = NULL, *cdata;
- size_t clen;
+ unsigned clen;
int ret;
unsigned long count;
struct page *page = (struct page *)(data);
}
/* reject if mean compression is too poor */
if ((clen > zv_max_mean_zsize) && (curr_pers_pampd_count > 0)) {
- total_zsize = xv_get_total_size_bytes(cli->xvpool);
+ total_zsize = zs_get_total_size_bytes(cli->zspool);
zv_mean_zsize = div_u64(total_zsize,
curr_pers_pampd_count);
if (zv_mean_zsize > zv_max_mean_zsize) {
goto out;
}
}
- pampd = (void *)zv_create(cli->xvpool, pool->pool_id,
+ pampd = (void *)zv_create(cli->zspool, pool->pool_id,
oid, index, cdata, clen);
if (pampd == NULL)
goto out;
atomic_dec(&zcache_curr_eph_pampd_count);
BUG_ON(atomic_read(&zcache_curr_eph_pampd_count) < 0);
} else {
- zv_free(cli->xvpool, (struct zv_hdr *)pampd);
+ zv_free(cli->zspool, pampd);
atomic_dec(&zcache_curr_pers_pampd_count);
BUG_ON(atomic_read(&zcache_curr_pers_pampd_count) < 0);
}
* zcache compression/decompression and related per-cpu stuff
*/
-#define LZO_WORKMEM_BYTES LZO1X_1_MEM_COMPRESS
-#define LZO_DSTMEM_PAGE_ORDER 1
-static DEFINE_PER_CPU(unsigned char *, zcache_workmem);
static DEFINE_PER_CPU(unsigned char *, zcache_dstmem);
+#define ZCACHE_DSTMEM_ORDER 1
-static int zcache_compress(struct page *from, void **out_va, size_t *out_len)
+static int zcache_compress(struct page *from, void **out_va, unsigned *out_len)
{
int ret = 0;
unsigned char *dmem = __get_cpu_var(zcache_dstmem);
- unsigned char *wmem = __get_cpu_var(zcache_workmem);
char *from_va;
BUG_ON(!irqs_disabled());
- if (unlikely(dmem == NULL || wmem == NULL))
- goto out; /* no buffer, so can't compress */
+ if (unlikely(dmem == NULL))
+ goto out; /* no buffer or no compressor so can't compress */
+ *out_len = PAGE_SIZE << ZCACHE_DSTMEM_ORDER;
- from_va = kmap_atomic(from, KM_USER0);
+ from_va = kmap_atomic(from);
mb();
- ret = lzo1x_1_compress(from_va, PAGE_SIZE, dmem, out_len, wmem);
- BUG_ON(ret != LZO_E_OK);
+ ret = zcache_comp_op(ZCACHE_COMPOP_COMPRESS, from_va, PAGE_SIZE, dmem,
+ out_len);
+ BUG_ON(ret);
*out_va = dmem;
- kunmap_atomic(from_va, KM_USER0);
+ kunmap_atomic(from_va);
ret = 1;
out:
return ret;
}
+static int zcache_comp_cpu_up(int cpu)
+{
+ struct crypto_comp *tfm;
+
+ tfm = crypto_alloc_comp(zcache_comp_name, 0, 0);
+ if (IS_ERR(tfm))
+ return NOTIFY_BAD;
+ *per_cpu_ptr(zcache_comp_pcpu_tfms, cpu) = tfm;
+ return NOTIFY_OK;
+}
+
+static void zcache_comp_cpu_down(int cpu)
+{
+ struct crypto_comp *tfm;
+
+ tfm = *per_cpu_ptr(zcache_comp_pcpu_tfms, cpu);
+ crypto_free_comp(tfm);
+ *per_cpu_ptr(zcache_comp_pcpu_tfms, cpu) = NULL;
+}
static int zcache_cpu_notifier(struct notifier_block *nb,
unsigned long action, void *pcpu)
{
- int cpu = (long)pcpu;
+ int ret, cpu = (long)pcpu;
struct zcache_preload *kp;
switch (action) {
case CPU_UP_PREPARE:
+ ret = zcache_comp_cpu_up(cpu);
+ if (ret != NOTIFY_OK) {
+ pr_err("zcache: can't allocate compressor transform\n");
+ return ret;
+ }
per_cpu(zcache_dstmem, cpu) = (void *)__get_free_pages(
- GFP_KERNEL | __GFP_REPEAT,
- LZO_DSTMEM_PAGE_ORDER),
- per_cpu(zcache_workmem, cpu) =
- kzalloc(LZO1X_MEM_COMPRESS,
- GFP_KERNEL | __GFP_REPEAT);
+ GFP_KERNEL | __GFP_REPEAT, ZCACHE_DSTMEM_ORDER);
break;
case CPU_DEAD:
case CPU_UP_CANCELED:
+ zcache_comp_cpu_down(cpu);
free_pages((unsigned long)per_cpu(zcache_dstmem, cpu),
- LZO_DSTMEM_PAGE_ORDER);
+ ZCACHE_DSTMEM_ORDER);
per_cpu(zcache_dstmem, cpu) = NULL;
- kfree(per_cpu(zcache_workmem, cpu));
- per_cpu(zcache_workmem, cpu) = NULL;
kp = &per_cpu(zcache_preloads, cpu);
while (kp->nr) {
kmem_cache_free(zcache_objnode_cache,
__setup("nofrontswap", no_frontswap);
+static int __init enable_zcache_compressor(char *s)
+{
+ strncpy(zcache_comp_name, s, ZCACHE_COMP_NAME_SZ);
+ zcache_enabled = 1;
+ return 1;
+}
+__setup("zcache=", enable_zcache_compressor);
+
+
+static int zcache_comp_init(void)
+{
+ int ret = 0;
+
+ /* check crypto algorithm */
+ if (*zcache_comp_name != '\0') {
+ ret = crypto_has_comp(zcache_comp_name, 0, 0);
+ if (!ret)
+ pr_info("zcache: %s not supported\n",
+ zcache_comp_name);
+ }
+ if (!ret)
+ strcpy(zcache_comp_name, "lzo");
+ ret = crypto_has_comp(zcache_comp_name, 0, 0);
+ if (!ret) {
+ ret = 1;
+ goto out;
+ }
+ pr_info("zcache: using %s compressor\n", zcache_comp_name);
+
+ /* alloc percpu transforms */
+ ret = 0;
+ zcache_comp_pcpu_tfms = alloc_percpu(struct crypto_comp *);
+ if (!zcache_comp_pcpu_tfms)
+ ret = 1;
+out:
+ return ret;
+}
+
static int __init zcache_init(void)
{
int ret = 0;
pr_err("zcache: can't register cpu notifier\n");
goto out;
}
+ ret = zcache_comp_init();
+ if (ret) {
+ pr_err("zcache: compressor initialization failed\n");
+ goto out;
+ }
for_each_online_cpu(cpu) {
void *pcpu = (void *)(long)cpu;
zcache_cpu_notifier(&zcache_cpu_notifier_block,
old_ops = zcache_frontswap_register_ops();
pr_info("zcache: frontswap enabled using kernel "
- "transcendent memory and xvmalloc\n");
+ "transcendent memory and zsmalloc\n");
if (old_ops.init != NULL)
pr_warning("zcache: frontswap_ops overridden");
}
struct zram *zram_devices;
/* Module params (documentation at end) */
-unsigned int zram_num_devices;
+static unsigned int num_devices;
static void zram_stat_inc(u32 *v)
{
static void zram_free_page(struct zram *zram, size_t index)
{
- u32 clen;
- void *obj;
+ void *handle = zram->table[index].handle;
- struct page *page = zram->table[index].page;
- u32 offset = zram->table[index].offset;
-
- if (unlikely(!page)) {
+ if (unlikely(!handle)) {
/*
* No memory is allocated for zero filled pages.
* Simply clear zero page flag.
}
if (unlikely(zram_test_flag(zram, index, ZRAM_UNCOMPRESSED))) {
- clen = PAGE_SIZE;
- __free_page(page);
+ __free_page(handle);
zram_clear_flag(zram, index, ZRAM_UNCOMPRESSED);
zram_stat_dec(&zram->stats.pages_expand);
goto out;
}
- obj = kmap_atomic(page) + offset;
- clen = xv_get_object_size(obj) - sizeof(struct zobj_header);
- kunmap_atomic(obj);
+ zs_free(zram->mem_pool, handle);
- xv_free(zram->mem_pool, page, offset);
- if (clen <= PAGE_SIZE / 2)
+ if (zram->table[index].size <= PAGE_SIZE / 2)
zram_stat_dec(&zram->stats.good_compress);
out:
- zram_stat64_sub(zram, &zram->stats.compr_size, clen);
+ zram_stat64_sub(zram, &zram->stats.compr_size,
+ zram->table[index].size);
zram_stat_dec(&zram->stats.pages_stored);
- zram->table[index].page = NULL;
- zram->table[index].offset = 0;
+ zram->table[index].handle = NULL;
+ zram->table[index].size = 0;
}
static void handle_zero_page(struct bio_vec *bvec)
struct page *page = bvec->bv_page;
void *user_mem;
- user_mem = kmap_atomic(page, KM_USER0);
+ user_mem = kmap_atomic(page);
memset(user_mem + bvec->bv_offset, 0, bvec->bv_len);
- kunmap_atomic(user_mem, KM_USER0);
+ kunmap_atomic(user_mem);
flush_dcache_page(page);
}
struct page *page = bvec->bv_page;
unsigned char *user_mem, *cmem;
- user_mem = kmap_atomic(page, KM_USER0);
- cmem = kmap_atomic(zram->table[index].handle, KM_USER1);
+ user_mem = kmap_atomic(page);
- cmem = kmap_atomic(zram->table[index].page);
++ cmem = kmap_atomic(zram->table[index].handle);
memcpy(user_mem + bvec->bv_offset, cmem + offset, bvec->bv_len);
- kunmap_atomic(cmem, KM_USER1);
- kunmap_atomic(user_mem, KM_USER0);
+ kunmap_atomic(cmem);
+ kunmap_atomic(user_mem);
flush_dcache_page(page);
}
}
/* Requested page is not present in compressed area */
- if (unlikely(!zram->table[index].page)) {
+ if (unlikely(!zram->table[index].handle)) {
pr_debug("Read before write: sector=%lu, size=%u",
(ulong)(bio->bi_sector), bio->bi_size);
handle_zero_page(bvec);
}
}
- user_mem = kmap_atomic(page, KM_USER0);
+ user_mem = kmap_atomic(page);
if (!is_partial_io(bvec))
uncmem = user_mem;
clen = PAGE_SIZE;
- cmem = kmap_atomic(zram->table[index].page) +
- zram->table[index].offset;
+ cmem = zs_map_object(zram->mem_pool, zram->table[index].handle);
ret = lzo1x_decompress_safe(cmem + sizeof(*zheader),
- xv_get_object_size(cmem) - sizeof(*zheader),
+ zram->table[index].size,
uncmem, &clen);
if (is_partial_io(bvec)) {
kfree(uncmem);
}
- kunmap_atomic(cmem);
+ zs_unmap_object(zram->mem_pool, zram->table[index].handle);
- kunmap_atomic(user_mem, KM_USER0);
+ kunmap_atomic(user_mem);
/* Should NEVER happen. Return bio error if it does. */
if (unlikely(ret != LZO_E_OK)) {
unsigned char *cmem;
if (zram_test_flag(zram, index, ZRAM_ZERO) ||
- !zram->table[index].page) {
+ !zram->table[index].handle) {
memset(mem, 0, PAGE_SIZE);
return 0;
}
- cmem = kmap_atomic(zram->table[index].page) +
- zram->table[index].offset;
+ cmem = zs_map_object(zram->mem_pool, zram->table[index].handle);
/* Page is stored uncompressed since it's incompressible */
if (unlikely(zram_test_flag(zram, index, ZRAM_UNCOMPRESSED))) {
memcpy(mem, cmem, PAGE_SIZE);
- kunmap_atomic(cmem, KM_USER0);
+ kunmap_atomic(cmem);
return 0;
}
ret = lzo1x_decompress_safe(cmem + sizeof(*zheader),
- xv_get_object_size(cmem) - sizeof(*zheader),
+ zram->table[index].size,
mem, &clen);
- kunmap_atomic(cmem);
+ zs_unmap_object(zram->mem_pool, zram->table[index].handle);
/* Should NEVER happen. Return bio error if it does. */
if (unlikely(ret != LZO_E_OK)) {
int ret;
u32 store_offset;
size_t clen;
+ void *handle;
struct zobj_header *zheader;
struct page *page, *page_store;
unsigned char *user_mem, *cmem, *src, *uncmem = NULL;
* System overwrites unused sectors. Free memory associated
* with this sector now.
*/
- if (zram->table[index].page ||
+ if (zram->table[index].handle ||
zram_test_flag(zram, index, ZRAM_ZERO))
zram_free_page(zram, index);
- user_mem = kmap_atomic(page, KM_USER0);
+ user_mem = kmap_atomic(page);
if (is_partial_io(bvec))
memcpy(uncmem + offset, user_mem + bvec->bv_offset,
uncmem = user_mem;
if (page_zero_filled(uncmem)) {
- kunmap_atomic(user_mem, KM_USER0);
+ kunmap_atomic(user_mem);
if (is_partial_io(bvec))
kfree(uncmem);
zram_stat_inc(&zram->stats.pages_zero);
ret = lzo1x_1_compress(uncmem, PAGE_SIZE, src, &clen,
zram->compress_workmem);
- kunmap_atomic(user_mem, KM_USER0);
+ kunmap_atomic(user_mem);
if (is_partial_io(bvec))
kfree(uncmem);
store_offset = 0;
zram_set_flag(zram, index, ZRAM_UNCOMPRESSED);
zram_stat_inc(&zram->stats.pages_expand);
- zram->table[index].page = page_store;
+ handle = page_store;
- src = kmap_atomic(page, KM_USER0);
- cmem = kmap_atomic(page_store, KM_USER1);
+ src = kmap_atomic(page);
++ cmem = kmap_atomic(page_store);
goto memstore;
}
- if (xv_malloc(zram->mem_pool, clen + sizeof(*zheader),
- &zram->table[index].page, &store_offset,
- GFP_NOIO | __GFP_HIGHMEM)) {
+ handle = zs_malloc(zram->mem_pool, clen + sizeof(*zheader));
+ if (!handle) {
pr_info("Error allocating memory for compressed "
"page: %u, size=%zu\n", index, clen);
ret = -ENOMEM;
goto out;
}
+ cmem = zs_map_object(zram->mem_pool, handle);
memstore:
- zram->table[index].offset = store_offset;
-
- cmem = kmap_atomic(zram->table[index].page) +
- zram->table[index].offset;
-
#if 0
/* Back-reference needed for memory defragmentation */
if (!zram_test_flag(zram, index, ZRAM_UNCOMPRESSED)) {
memcpy(cmem, src, clen);
- kunmap_atomic(cmem);
- if (unlikely(zram_test_flag(zram, index, ZRAM_UNCOMPRESSED)))
+ if (unlikely(zram_test_flag(zram, index, ZRAM_UNCOMPRESSED))) {
- kunmap_atomic(cmem, KM_USER1);
- kunmap_atomic(src, KM_USER0);
++ kunmap_atomic(cmem);
+ kunmap_atomic(src);
+ } else {
+ zs_unmap_object(zram->mem_pool, handle);
+ }
+
+ zram->table[index].handle = handle;
+ zram->table[index].size = clen;
/* Update stats */
zram_stat64_add(zram, &zram->stats.compr_size, clen);
/* Free all pages that are still in this zram device */
for (index = 0; index < zram->disksize >> PAGE_SHIFT; index++) {
- struct page *page;
- u16 offset;
-
- page = zram->table[index].page;
- offset = zram->table[index].offset;
-
- if (!page)
+ void *handle = zram->table[index].handle;
+ if (!handle)
continue;
if (unlikely(zram_test_flag(zram, index, ZRAM_UNCOMPRESSED)))
- __free_page(page);
+ __free_page(handle);
else
- xv_free(zram->mem_pool, page, offset);
+ zs_free(zram->mem_pool, handle);
}
vfree(zram->table);
zram->table = NULL;
- xv_destroy_pool(zram->mem_pool);
+ zs_destroy_pool(zram->mem_pool);
zram->mem_pool = NULL;
/* Reset stats */
/* zram devices sort of resembles non-rotational disks */
queue_flag_set_unlocked(QUEUE_FLAG_NONROT, zram->disk->queue);
- zram->mem_pool = xv_create_pool();
+ zram->mem_pool = zs_create_pool("zram", GFP_NOIO | __GFP_HIGHMEM);
if (!zram->mem_pool) {
pr_err("Error creating memory pool\n");
ret = -ENOMEM;
blk_cleanup_queue(zram->queue);
}
+unsigned int zram_get_num_devices(void)
+{
+ return num_devices;
+}
+
static int __init zram_init(void)
{
int ret, dev_id;
- if (zram_num_devices > max_num_devices) {
+ if (num_devices > max_num_devices) {
pr_warning("Invalid value for num_devices: %u\n",
- zram_num_devices);
+ num_devices);
ret = -EINVAL;
goto out;
}
goto out;
}
- if (!zram_num_devices) {
+ if (!num_devices) {
pr_info("num_devices not specified. Using default: 1\n");
- zram_num_devices = 1;
+ num_devices = 1;
}
/* Allocate the device array and initialize each one */
- pr_info("Creating %u devices ...\n", zram_num_devices);
- zram_devices = kzalloc(zram_num_devices * sizeof(struct zram), GFP_KERNEL);
+ pr_info("Creating %u devices ...\n", num_devices);
+ zram_devices = kzalloc(num_devices * sizeof(struct zram), GFP_KERNEL);
if (!zram_devices) {
ret = -ENOMEM;
goto unregister;
}
- for (dev_id = 0; dev_id < zram_num_devices; dev_id++) {
+ for (dev_id = 0; dev_id < num_devices; dev_id++) {
ret = create_device(&zram_devices[dev_id], dev_id);
if (ret)
goto free_devices;
int i;
struct zram *zram;
- for (i = 0; i < zram_num_devices; i++) {
+ for (i = 0; i < num_devices; i++) {
zram = &zram_devices[i];
destroy_device(zram);
pr_debug("Cleanup done!\n");
}
-module_param(zram_num_devices, uint, 0);
-MODULE_PARM_DESC(zram_num_devices, "Number of zram devices");
+module_param(num_devices, uint, 0);
+MODULE_PARM_DESC(num_devices, "Number of zram devices");
module_init(zram_init);
module_exit(zram_exit);
#include <trace/events/task.h>
#include "internal.h"
+#include <trace/events/sched.h>
+
int core_uses_pid;
char core_pattern[CORENAME_MAX_SIZE] = "core";
unsigned int core_pipe_limit;
if (old_mm) {
up_read(&old_mm->mmap_sem);
BUG_ON(active_mm != old_mm);
+ setmax_mm_hiwater_rss(&tsk->signal->maxrss, old_mm);
mm_update_next_owner(old_mm);
mmput(old_mm);
return 0;
sig->notify_count = 0;
no_thread_group:
- if (current->mm)
- setmax_mm_hiwater_rss(&sig->maxrss, current->mm);
+ /* we have changed execution domain */
+ tsk->exit_signal = SIGCHLD;
exit_itimers(sig);
flush_itimer_signals();
ret = -EFAULT;
goto out;
}
- kaddr = kmap_atomic(page, KM_USER0);
+ kaddr = kmap_atomic(page);
for (; offset < PAGE_SIZE && kaddr[offset];
offset++, bprm->p++)
;
- kunmap_atomic(kaddr, KM_USER0);
+ kunmap_atomic(kaddr);
put_arg_page(page);
if (offset == PAGE_SIZE)
*/
bprm->recursion_depth = depth;
if (retval >= 0) {
- if (depth == 0)
- ptrace_event(PTRACE_EVENT_EXEC,
- old_pid);
+ if (depth == 0) {
+ trace_sched_process_exec(current, old_pid, bprm);
+ ptrace_event(PTRACE_EVENT_EXEC, old_pid);
+ }
put_binfmt(fmt);
allow_write_access(bprm->file);
if (bprm->file)
return 1;
}
+/*
+ * We can do the critical dentry name comparison and hashing
+ * operations one word at a time, but we are limited to:
+ *
+ * - Architectures with fast unaligned word accesses. We could
+ * do a "get_unaligned()" if this helps and is sufficiently
+ * fast.
+ *
+ * - Little-endian machines (so that we can generate the mask
+ * of low bytes efficiently). Again, we *could* do a byte
+ * swapping load on big-endian architectures if that is not
+ * expensive enough to make the optimization worthless.
+ *
+ * - non-CONFIG_DEBUG_PAGEALLOC configurations (so that we
+ * do not trap on the (extremely unlikely) case of a page
+ * crossing operation.
+ *
+ * - Furthermore, we need an efficient 64-bit compile for the
+ * 64-bit case in order to generate the "number of bytes in
+ * the final mask". Again, that could be replaced with a
+ * efficient population count instruction or similar.
+ */
+#ifdef CONFIG_DCACHE_WORD_ACCESS
+
+#ifdef CONFIG_64BIT
+
+/*
+ * Jan Achrenius on G+: microoptimized version of
+ * the simpler "(mask & ONEBYTES) * ONEBYTES >> 56"
+ * that works for the bytemasks without having to
+ * mask them first.
+ */
+static inline long count_masked_bytes(unsigned long mask)
+{
+ return mask*0x0001020304050608 >> 56;
+}
+
+static inline unsigned int fold_hash(unsigned long hash)
+{
+ hash += hash >> (8*sizeof(int));
+ return hash;
+}
+
+#else /* 32-bit case */
+
+/* Carl Chatfield / Jan Achrenius G+ version for 32-bit */
+static inline long count_masked_bytes(long mask)
+{
+ /* (000000 0000ff 00ffff ffffff) -> ( 1 1 2 3 ) */
+ long a = (0x0ff0001+mask) >> 23;
+ /* Fix the 1 for 00 case */
+ return a & mask;
+}
+
+#define fold_hash(x) (x)
+
+#endif
+
+unsigned int full_name_hash(const unsigned char *name, unsigned int len)
+{
+ unsigned long a, mask;
+ unsigned long hash = 0;
+
+ for (;;) {
+ a = *(unsigned long *)name;
+ hash *= 9;
+ if (len < sizeof(unsigned long))
+ break;
+ hash += a;
+ name += sizeof(unsigned long);
+ len -= sizeof(unsigned long);
+ if (!len)
+ goto done;
+ }
+ mask = ~(~0ul << len*8);
+ hash += mask & a;
+done:
+ return fold_hash(hash);
+}
+EXPORT_SYMBOL(full_name_hash);
+
+#define ONEBYTES 0x0101010101010101ul
+#define SLASHBYTES 0x2f2f2f2f2f2f2f2ful
+#define HIGHBITS 0x8080808080808080ul
+
+/* Return the high bit set in the first byte that is a zero */
+static inline unsigned long has_zero(unsigned long a)
+{
+ return ((a - ONEBYTES) & ~a) & HIGHBITS;
+}
+
+/*
+ * Calculate the length and hash of the path component, and
+ * return the length of the component;
+ */
+static inline unsigned long hash_name(const char *name, unsigned int *hashp)
+{
+ unsigned long a, mask, hash, len;
+
+ hash = a = 0;
+ len = -sizeof(unsigned long);
+ do {
+ hash = (hash + a) * 9;
+ len += sizeof(unsigned long);
+ a = *(unsigned long *)(name+len);
+ /* Do we have any NUL or '/' bytes in this word? */
+ mask = has_zero(a) | has_zero(a ^ SLASHBYTES);
+ } while (!mask);
+
+ /* The mask *below* the first high bit set */
+ mask = (mask - 1) & ~mask;
+ mask >>= 7;
+ hash += a & mask;
+ *hashp = fold_hash(hash);
+
+ return len + count_masked_bytes(mask);
+}
+
+#else
+
unsigned int full_name_hash(const unsigned char *name, unsigned int len)
{
unsigned long hash = init_name_hash();
return len;
}
+#endif
+
/*
* Name resolution.
* This is the basic name resolution function, turning a pathname into
if (err)
goto fail;
- kaddr = kmap_atomic(page, KM_USER0);
+ kaddr = kmap_atomic(page);
memcpy(kaddr, symname, len-1);
- kunmap_atomic(kaddr, KM_USER0);
+ kunmap_atomic(kaddr);
err = pagecache_write_end(NULL, mapping, 0, len-1, len-1,
page, fsdata);
to_copy = min(RDS_FRAG_SIZE - frag_off, PAGE_SIZE - map_off);
BUG_ON(to_copy & 7); /* Must be 64bit aligned. */
- addr = kmap_atomic(sg_page(&frag->f_sg), KM_SOFTIRQ0);
+ addr = kmap_atomic(sg_page(&frag->f_sg));
src = addr + frag_off;
dst = (void *)map->m_page_addrs[map_page] + map_off;
uncongested |= ~(*src) & *dst;
*dst++ = *src++;
}
- kunmap_atomic(addr, KM_SOFTIRQ0);
+ kunmap_atomic(addr);
copied += to_copy;
if (data_len < sizeof(struct rds_header)) {
rds_ib_conn_error(conn, "incoming message "
- "from %pI4 didn't inclue a "
+ "from %pI4 didn't include a "
"header, disconnecting and "
"reconnecting\n",
&conn->c_faddr);
rds_ib_cong_recv(conn, ibinc);
else {
rds_recv_incoming(conn, conn->c_faddr, conn->c_laddr,
- &ibinc->ii_inc, GFP_ATOMIC,
- KM_SOFTIRQ0);
+ &ibinc->ii_inc, GFP_ATOMIC);
state->ack_next = be64_to_cpu(hdr->h_sequence);
state->ack_next_valid = 1;
}
to_copy = min(RDS_FRAG_SIZE - frag_off, PAGE_SIZE - map_off);
BUG_ON(to_copy & 7); /* Must be 64bit aligned. */
- addr = kmap_atomic(frag->f_page, KM_SOFTIRQ0);
+ addr = kmap_atomic(frag->f_page);
src = addr + frag_off;
dst = (void *)map->m_page_addrs[map_page] + map_off;
uncongested |= ~(*src) & *dst;
*dst++ = *src++;
}
- kunmap_atomic(addr, KM_SOFTIRQ0);
+ kunmap_atomic(addr);
copied += to_copy;
if (byte_len < sizeof(struct rds_header)) {
rds_iw_conn_error(conn, "incoming message "
- "from %pI4 didn't inclue a "
+ "from %pI4 didn't include a "
"header, disconnecting and "
"reconnecting\n",
&conn->c_faddr);
rds_iw_cong_recv(conn, iwinc);
else {
rds_recv_incoming(conn, conn->c_faddr, conn->c_laddr,
- &iwinc->ii_inc, GFP_ATOMIC,
- KM_SOFTIRQ0);
+ &iwinc->ii_inc, GFP_ATOMIC);
state->ack_next = be64_to_cpu(hdr->h_sequence);
state->ack_next_valid = 1;
}
projects / platform / adaptation / renesas_rcar / renesas_kernel.git / commitdiff