platform/kernel/linux-rpi.git
3 years agoKVM: PPC: Book3S HV P9: Restore host CTRL SPR after guest exit
Nicholas Piggin [Mon, 12 Apr 2021 01:48:34 +0000 (11:48 +1000)]
KVM: PPC: Book3S HV P9: Restore host CTRL SPR after guest exit

[ Upstream commit 5088eb4092df12d701af8e0e92860b7186365279 ]

The host CTRL (runlatch) value is not restored after guest exit. The
host CTRL should always be 1 except in CPU idle code, so this can result
in the host running with runlatch clear, and potentially switching to
a different vCPU which then runs with runlatch clear as well.

This has little effect on P9 machines, CTRL is only responsible for some
PMU counter logic in the host and so other than corner cases of software
relying on that, or explicitly reading the runlatch value (Linux does
not appear to be affected but it's possible non-Linux guests could be),
there should be no execution correctness problem, though it could be
used as a covert channel between guests.

There may be microcontrollers, firmware or monitoring tools that sample
the runlatch value out-of-band, however since the register is writable
by guests, these values would (should) not be relied upon for correct
operation of the host, so suboptimal performance or incorrect reporting
should be the worst problem.

Fixes: 95a6432ce9038 ("KVM: PPC: Book3S HV: Streamlined guest entry/exit path on P9 for radix guests")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210412014845.1517916-2-npiggin@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7663s: fix the possible device hang in high traffic
Sean Wang [Thu, 4 Mar 2021 13:25:23 +0000 (21:25 +0800)]
mt76: mt7663s: fix the possible device hang in high traffic

[ Upstream commit 45247a85614b49b07b9dc59a4e6783b17e766ff2 ]

Use the additional memory barrier to ensure the skb list up-to-date
between the skb producer and consumer to avoid the invalid skb content
written into sdio controller and then cause device hang due to mcu assert
caught by WR_TIMEOUT_INT.

Fixes: 1522ff731f84 ("mt76: mt7663s: introduce sdio tx aggregation")
Signed-off-by: Sean Wang <sean.wang@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7663s: make all of packets 4-bytes aligned in sdio tx aggregation
Sean Wang [Thu, 4 Mar 2021 13:25:22 +0000 (21:25 +0800)]
mt76: mt7663s: make all of packets 4-bytes aligned in sdio tx aggregation

[ Upstream commit 455ae5aabcc72fed7e5c803d59d122415500dc08 ]

Each packet should be padded with the additional zero to become 4-bytes
alignment in sdio tx aggregation.

Fixes: 1522ff731f84 ("mt76: mt7663s: introduce sdio tx aggregation")
Signed-off-by: Sean Wang <sean.wang@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7915: fix mib stats counter reporting to mac80211
Ryder Lee [Tue, 23 Feb 2021 14:00:36 +0000 (22:00 +0800)]
mt76: mt7915: fix mib stats counter reporting to mac80211

[ Upstream commit 2b35050a321865859fd2f12a3c18ed7be27858c9 ]

In order to properly report MIB counters to mac80211, resets stats in
mt7915_get_stats routine() and hold mt76 mutex accessing MIB counters.
Sum up MIB counters in mt7915_mac_update_mib_stats routine.

Fixes: e57b7901469f ("mt76: add mac80211 driver for MT7915 PCIe-based chipsets")
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7615: fix mib stats counter reporting to mac80211
Lorenzo Bianconi [Tue, 23 Feb 2021 10:42:49 +0000 (11:42 +0100)]
mt76: mt7615: fix mib stats counter reporting to mac80211

[ Upstream commit 2eb6f6c437745bce46bd7a8f3a22a732d5b9becb ]

In order to properly report MIB counters to mac80211, resets stats in
mt7615_get_stats routine and hold mt76 mutex accessing MIB counters.
Sum up MIB counters in mt7615_mac_update_mib_stats routine.

Fixes: c388d8584bc83 ("mt76: mt7615: add a get_stats() callback")
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7915: fix aggr len debugfs node
Lorenzo Bianconi [Mon, 22 Feb 2021 01:14:36 +0000 (02:14 +0100)]
mt76: mt7915: fix aggr len debugfs node

[ Upstream commit 9fb9d755fae20b5ad62ef8b4e9289e5baea2c6fc ]

Similar to mt7921, fix 802.11 aggr len debugfs reporting for mt7915 driver.

Fixes: e57b7901469fc ("mt76: add mac80211 driver for MT7915 PCIe-based chipsets")
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7915: fix tx skb dma unmap
Felix Fietkau [Sat, 13 Feb 2021 17:44:10 +0000 (18:44 +0100)]
mt76: mt7915: fix tx skb dma unmap

[ Upstream commit 7dcf3c04f0aca746517a77433b33d40868ca4749 ]

The first pointer in the txp needs to be unmapped as well, otherwise it will
leak DMA mapping entries

Reported-by: Ben Greear <greearb@candelatech.com>
Fixes: 27d5c528a7ca ("mt76: fix double DMA unmap of the first buffer on 7615/7915")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt76: mt7615: fix tx skb dma unmap
Felix Fietkau [Sat, 13 Feb 2021 17:42:15 +0000 (18:42 +0100)]
mt76: mt7615: fix tx skb dma unmap

[ Upstream commit ebee7885bb12a8fe2c2f9bac87dbd87a05b645f9 ]

The first pointer in the txp needs to be unmapped as well, otherwise it will
leak DMA mapping entries

Fixes: 27d5c528a7ca ("mt76: fix double DMA unmap of the first buffer on 7615/7915")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomt7601u: fix always true expression
Colin Ian King [Thu, 25 Feb 2021 18:32:41 +0000 (18:32 +0000)]
mt7601u: fix always true expression

[ Upstream commit 87fce88658ba047ae62e83497d3f3c5dc22fa6f9 ]

Currently the expression ~nic_conf1 is always true because nic_conf1
is a u16 and according to 6.5.3.3 of the C standard the ~ operator
promotes the u16 to an integer before flipping all the bits. Thus
the top 16 bits of the integer result are all set so the expression
is always true.  If the intention was to flip all the bits of nic_conf1
then casting the integer result back to a u16 is a suitabel fix.

Interestingly static analyzers seem to thing a bitwise ! should be
used instead of ~ for this scenario, so I think the original intent
of the expression may need some extra consideration.

Addresses-Coverity: ("Logical vs. bitwise operator")
Fixes: c869f77d6abb ("add mt7601u driver")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Jakub Kicinski <kubakici@wp.pl>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20210225183241.1002129-1-colin.king@canonical.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agortw88: Fix an error code in rtw_debugfs_set_rsvd_page()
Dan Carpenter [Mon, 15 Mar 2021 09:23:37 +0000 (12:23 +0300)]
rtw88: Fix an error code in rtw_debugfs_set_rsvd_page()

[ Upstream commit c9eaee0c2ec6b1002044fb698cdfb5d9ef4ed28c ]

The sscanf() function returns the number of matches (0 or 1 in this
case).  It doesn't return error codes.  We should return -EINVAL if the
string is invalid

Fixes: c376c1fc87b7 ("rtw88: add h2c command in debugfs")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/YE8nmatMDBDDWkjq@mwanda
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoxfs: fix return of uninitialized value in variable error
Colin Ian King [Fri, 9 Apr 2021 17:27:34 +0000 (10:27 -0700)]
xfs: fix return of uninitialized value in variable error

[ Upstream commit 3b6dd9a9aeeada19d0c820ff68e979243a888bb6 ]

A previous commit removed a call to xfs_attr3_leaf_read that
assigned an error return code to variable error. We now have
a few early error return paths to label 'out' that return
error if error is set; however error now is uninitialized
so potentially garbage is being returned.  Fix this by setting
error to zero to restore the original behaviour where error
was zero at the label 'restart'.

Addresses-Coverity: ("Uninitialized scalar variable")
Fixes: 07120f1abdff ("xfs: Add xfs_has_attr and subroutines")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoperf vendor events amd: Fix broken L2 Cache Hits from L2 HWPF metric
Smita Koralahalli [Tue, 6 Apr 2021 21:59:41 +0000 (16:59 -0500)]
perf vendor events amd: Fix broken L2 Cache Hits from L2 HWPF metric

[ Upstream commit 86c2bc3da769124e3e856b6e9457be3667c30919 ]

Commit 08ed77e414ab2342 ("perf vendor events amd: Add recommended events")
added the hits event "L2 Cache Hits from L2 HWPF" with the same metric
expression as the accesses event "L2 Cache Accesses from L2 HWPF":

$ perf list --details
...
  l2_cache_accesses_from_l2_hwpf
     [L2 Cache Accesses from L2 HWPF]
     [l2_pf_hit_l2 + l2_pf_miss_l2_hit_l3 + l2_pf_miss_l2_l3]
  l2_cache_hits_from_l2_hwpf
     [L2 Cache Hits from L2 HWPF]
     [l2_pf_hit_l2 + l2_pf_miss_l2_hit_l3 + l2_pf_miss_l2_l3]
...

This was wrong and led to counting hits the same as accesses. Section
2.1.15.2 "Performance Measurement" of "PPR for AMD Family 17h Model 31h
B0 - 55803 Rev 0.54 - Sep 12, 2019", documents the hits event with
EventCode 0x70 which is the same as l2_pf_hit_l2.

Fix this, and massage the description for l2_pf_hit_l2 as the hits event
is now the duplicate of l2_pf_hit_l2. AMD recommends using the recommended
event over other events if the duplicate exists and maintain both for
consistency. Hence, l2_cache_hits_from_l2_hwpf should override
l2_pf_hit_l2.

Before:

 # perf stat -M l2_cache_accesses_from_l2_hwpf,l2_cache_hits_from_l2_hwpf sleep 1

 Performance counter stats for 'sleep 1':

             1,436      l2_pf_miss_l2_l3          # 11114.00 l2_cache_accesses_from_l2_hwpf
                                                  # 11114.00 l2_cache_hits_from_l2_hwpf
             4,482      l2_pf_hit_l2
             5,196      l2_pf_miss_l2_hit_l3

       1.001765339 seconds time elapsed

After:

 # perf stat -M l2_cache_accesses_from_l2_hwpf sleep 1

 Performance counter stats for 'sleep 1':

             1,477      l2_pf_miss_l2_l3          # 10442.00 l2_cache_accesses_from_l2_hwpf
             3,978      l2_pf_hit_l2
             4,987      l2_pf_miss_l2_hit_l3

       1.001491186 seconds time elapsed

 # perf stat -e l2_cache_hits_from_l2_hwpf sleep 1

 Performance counter stats for 'sleep 1':

             3,983      l2_cache_hits_from_l2_hwpf

       1.001329970 seconds time elapsed

Note the difference in performance counter values for the accesses
versus the hits after the fix, and the hits event now counting the same
as l2_pf_hit_l2.

Fixes: 08ed77e414ab ("perf vendor events amd: Add recommended events")
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=206537
Reviewed-by: Robert Richter <rrichter@amd.com>
Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com>
Tested-by: Arnaldo Carvalho de Melo <acme@kernel.org> # On a 3900X
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kim Phillips <kim.phillips@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Martin Liška <mliska@suse.cz>
Cc: Michael Petlan <mpetlan@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Vijay Thakkar <vijaythakkar@me.com>
Cc: linux-perf-users@vger.kernel.org
Link: https://lore.kernel.org/r/20210406215944.113332-2-Smita.KoralahalliChannabasappa@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomac80211: bail out if cipher schemes are invalid
Johannes Berg [Thu, 8 Apr 2021 12:31:50 +0000 (14:31 +0200)]
mac80211: bail out if cipher schemes are invalid

[ Upstream commit db878e27a98106a70315d264cc92230d84009e72 ]

If any of the cipher schemes specified by the driver are invalid, bail
out and fail the registration rather than just warning.  Otherwise, we
might later crash when we try to use the invalid cipher scheme, e.g.
if the hdr_len is (significantly) less than the pn_offs + pn_len, we'd
have an out-of-bounds access in RX validation.

Fixes: 2475b1cc0d52 ("mac80211: add generic cipher scheme support")
Link: https://lore.kernel.org/r/20210408143149.38a3a13a1b19.I6b7f5790fa0958ed8049cf02ac2a535c61e9bc96@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc: iommu: fix build when neither PCI or IBMVIO is set
Randy Dunlap [Sun, 4 Apr 2021 19:26:23 +0000 (12:26 -0700)]
powerpc: iommu: fix build when neither PCI or IBMVIO is set

[ Upstream commit b27dadecdf9102838331b9a0b41ffc1cfe288154 ]

When neither CONFIG_PCI nor CONFIG_IBMVIO is set/enabled, iommu.c has a
build error. The fault injection code is not useful in that kernel config,
so make the FAIL_IOMMU option depend on PCI || IBMVIO.

Prevents this build error (warning escalated to error):
../arch/powerpc/kernel/iommu.c:178:30: error: 'fail_iommu_bus_notifier' defined but not used [-Werror=unused-variable]
  178 | static struct notifier_block fail_iommu_bus_notifier = {

Fixes: d6b9a81b2a45 ("powerpc: IOMMU fault injection")
Reported-by: kernel test robot <lkp@intel.com>
Suggested-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210404192623.10697-1-rdunlap@infradead.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc/perf: Fix PMU constraint check for EBB events
Athira Rajeev [Tue, 6 Apr 2021 16:16:01 +0000 (12:16 -0400)]
powerpc/perf: Fix PMU constraint check for EBB events

[ Upstream commit 10f8f96179ecc7f69c927f6d231f6d02736cea83 ]

The power PMU group constraints includes check for EBB events to make
sure all events in a group must agree on EBB. This will prevent
scheduling EBB and non-EBB events together. But in the existing check,
settings for constraint mask and value is interchanged. Patch fixes the
same.

Before the patch, PMU selftest "cpu_event_pinned_vs_ebb_test" fails with
below in dmesg logs. This happens because EBB event gets enabled along
with a non-EBB cpu event.

  [35600.453346] cpu_event_pinne[41326]: illegal instruction (4)
  at 10004a18 nip 10004a18 lr 100049f8 code 1 in
  cpu_event_pinned_vs_ebb_test[10000000+10000]

Test results after the patch:

  $ ./pmu/ebb/cpu_event_pinned_vs_ebb_test
  test: cpu_event_pinned_vs_ebb
  tags: git_version:v5.12-rc5-93-gf28c3125acd3-dirty
  Binding to cpu 8
  EBB Handler is at 0x100050c8
  read error on event 0x7fffe6bd4040!
  PM_RUN_INST_CMPL: result 9872 running/enabled 37930432
  success: cpu_event_pinned_vs_ebb

This bug was hidden by other logic until commit 1908dc911792 (perf:
Tweak perf_event_attr::exclusive semantics).

Fixes: 4df489991182 ("powerpc/perf: Add power8 EBB support")
Reported-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com>
Signed-off-by: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
[mpe: Mention commit 1908dc911792]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/1617725761-1464-1-git-send-email-atrajeev@linux.vnet.ibm.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc/64s: Fix pte update for kernel memory on radix
Jordan Niethe [Mon, 8 Feb 2021 03:29:56 +0000 (14:29 +1100)]
powerpc/64s: Fix pte update for kernel memory on radix

[ Upstream commit b8b2f37cf632434456182e9002d63cbc4cccc50c ]

When adding a PTE a ptesync is needed to order the update of the PTE
with subsequent accesses otherwise a spurious fault may be raised.

radix__set_pte_at() does not do this for performance gains. For
non-kernel memory this is not an issue as any faults of this kind are
corrected by the page fault handler. For kernel memory these faults
are not handled. The current solution is that there is a ptesync in
flush_cache_vmap() which should be called when mapping from the
vmalloc region.

However, map_kernel_page() does not call flush_cache_vmap(). This is
troublesome in particular for code patching with Strict RWX on radix.
In do_patch_instruction() the page frame that contains the instruction
to be patched is mapped and then immediately patched. With no ordering
or synchronization between setting up the PTE and writing to the page
it is possible for faults.

As the code patching is done using __put_user_asm_goto() the resulting
fault is obscured - but using a normal store instead it can be seen:

  BUG: Unable to handle kernel data access on write at 0xc008000008f24a3c
  Faulting instruction address: 0xc00000000008bd74
  Oops: Kernel access of bad area, sig: 11 [#1]
  LE PAGE_SIZE=64K MMU=Radix SMP NR_CPUS=2048 NUMA PowerNV
  Modules linked in: nop_module(PO+) [last unloaded: nop_module]
  CPU: 4 PID: 757 Comm: sh Tainted: P           O      5.10.0-rc5-01361-ge3c1b78c8440-dirty #43
  NIP:  c00000000008bd74 LR: c00000000008bd50 CTR: c000000000025810
  REGS: c000000016f634a0 TRAP: 0300   Tainted: P           O       (5.10.0-rc5-01361-ge3c1b78c8440-dirty)
  MSR:  9000000000009033 <SF,HV,EE,ME,IR,DR,RI,LE>  CR: 44002884  XER: 00000000
  CFAR: c00000000007c68c DAR: c008000008f24a3c DSISR: 42000000 IRQMASK: 1

This results in the kind of issue reported here:
  https://lore.kernel.org/linuxppc-dev/15AC5B0E-A221-4B8C-9039-FA96B8EF7C88@lca.pw/

Chris Riedl suggested a reliable way to reproduce the issue:
  $ mount -t debugfs none /sys/kernel/debug
  $ (while true; do echo function > /sys/kernel/debug/tracing/current_tracer ; echo nop > /sys/kernel/debug/tracing/current_tracer ; done) &

Turning ftrace on and off does a large amount of code patching which
in usually less then 5min will crash giving a trace like:

   ftrace-powerpc: (____ptrval____): replaced (4b473b11) != old (60000000)
   ------------[ ftrace bug ]------------
   ftrace failed to modify
   [<c000000000bf8e5c>] napi_busy_loop+0xc/0x390
    actual:   11:3b:47:4b
   Setting ftrace call site to call ftrace function
   ftrace record flags: 80000001
    (1)
    expected tramp: c00000000006c96c
   ------------[ cut here ]------------
   WARNING: CPU: 4 PID: 809 at kernel/trace/ftrace.c:2065 ftrace_bug+0x28c/0x2e8
   Modules linked in: nop_module(PO-) [last unloaded: nop_module]
   CPU: 4 PID: 809 Comm: sh Tainted: P           O      5.10.0-rc5-01360-gf878ccaf250a #1
   NIP:  c00000000024f334 LR: c00000000024f330 CTR: c0000000001a5af0
   REGS: c000000004c8b760 TRAP: 0700   Tainted: P           O       (5.10.0-rc5-01360-gf878ccaf250a)
   MSR:  900000000282b033 <SF,HV,VEC,VSX,EE,FP,ME,IR,DR,RI,LE>  CR: 28008848  XER: 20040000
   CFAR: c0000000001a9c98 IRQMASK: 0
   GPR00: c00000000024f330 c000000004c8b9f0 c000000002770600 0000000000000022
   GPR04: 00000000ffff7fff c000000004c8b6d0 0000000000000027 c0000007fe9bcdd8
   GPR08: 0000000000000023 ffffffffffffffd8 0000000000000027 c000000002613118
   GPR12: 0000000000008000 c0000007fffdca00 0000000000000000 0000000000000000
   GPR16: 0000000023ec37c5 0000000000000000 0000000000000000 0000000000000008
   GPR20: c000000004c8bc90 c0000000027a2d20 c000000004c8bcd0 c000000002612fe8
   GPR24: 0000000000000038 0000000000000030 0000000000000028 0000000000000020
   GPR28: c000000000ff1b68 c000000000bf8e5c c00000000312f700 c000000000fbb9b0
   NIP ftrace_bug+0x28c/0x2e8
   LR  ftrace_bug+0x288/0x2e8
   Call Trace:
     ftrace_bug+0x288/0x2e8 (unreliable)
     ftrace_modify_all_code+0x168/0x210
     arch_ftrace_update_code+0x18/0x30
     ftrace_run_update_code+0x44/0xc0
     ftrace_startup+0xf8/0x1c0
     register_ftrace_function+0x4c/0xc0
     function_trace_init+0x80/0xb0
     tracing_set_tracer+0x2a4/0x4f0
     tracing_set_trace_write+0xd4/0x130
     vfs_write+0xf0/0x330
     ksys_write+0x84/0x140
     system_call_exception+0x14c/0x230
     system_call_common+0xf0/0x27c

To fix this when updating kernel memory PTEs using ptesync.

Fixes: f1cb8f9beba8 ("powerpc/64s/radix: avoid ptesync after set_pte and ptep_set_access_flags")
Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
[mpe: Tidy up change log slightly]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210208032957.1232102-1-jniethe5@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoIB/hfi1: Use kzalloc() for mmu_rb_handler allocation
Mike Marciniszyn [Mon, 29 Mar 2021 13:54:14 +0000 (09:54 -0400)]
IB/hfi1: Use kzalloc() for mmu_rb_handler allocation

[ Upstream commit ca5f72568e034e1295a7ae350b1f786fcbfb2848 ]

The code currently assumes that the mmu_notifier struct
embedded in mmu_rb_handler only contains two fields.

There are now extra fields:

struct mmu_notifier {
        struct hlist_node hlist;
        const struct mmu_notifier_ops *ops;
        struct mm_struct *mm;
        struct rcu_head rcu;
        unsigned int users;
};

Given that there in no init for the mmu_notifier, a kzalloc() should
be used to insure that any newly added fields are given a predictable
initial value of zero.

Fixes: 06e0ffa69312 ("IB/hfi1: Re-factor MMU notification code")
Link: https://lore.kernel.org/r/1617026056-50483-9-git-send-email-dennis.dalessandro@cornelisnetworks.com
Reviewed-by: Adam Goldman <adam.goldman@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoliquidio: Fix unintented sign extension of a left shift of a u16
Colin Ian King [Wed, 7 Apr 2021 10:12:48 +0000 (11:12 +0100)]
liquidio: Fix unintented sign extension of a left shift of a u16

[ Upstream commit 298b58f00c0f86868ea717426beb5c1198772f81 ]

The macro CN23XX_PEM_BAR1_INDEX_REG is being used to shift oct->pcie_port
(a u16) left 24 places. There are two subtle issues here, first the
shift gets promoted to an signed int and then sign extended to a u64.
If oct->pcie_port is 0x80 or more then the upper bits get sign extended
to 1. Secondly shfiting a u16 24 bits will lead to an overflow so it
needs to be cast to a u64 for all the bits to not overflow.

It is entirely possible that the u16 port value is never large enough
for this to fail, but it is useful to fix unintended overflows such
as this.

Fix this by casting the port parameter to the macro to a u64 before
the shift.

Addresses-Coverity: ("Unintended sign extension")
Fixes: 5bc67f587ba7 ("liquidio: CN23XX register definitions")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoASoC: simple-card: fix possible uninitialized single_cpu local variable
Krzysztof Kozlowski [Wed, 7 Apr 2021 09:20:27 +0000 (11:20 +0200)]
ASoC: simple-card: fix possible uninitialized single_cpu local variable

[ Upstream commit fa74c223b6fd78a5314b4c61b9abdbed3c2185b4 ]

The 'single_cpu' local variable is assigned by asoc_simple_parse_dai()
and later used in a asoc_simple_canonicalize_cpu() call, assuming the
entire function did not exit on errors.

However the first function returns 0 if passed device_node is NULL,
thus leaving the variable uninitialized and reporting success.

Addresses-Coverity: Uninitialized scalar variable
Fixes: 8f7f298a3337 ("ASoC: simple-card-utils: separate asoc_simple_card_parse_dai()")
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Acked-by: Sameer Pujar <spujar@nvidia.com>
Link: https://lore.kernel.org/r/20210407092027.60769-1-krzysztof.kozlowski@canonical.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoKVM: arm64: Initialize VCPU mdcr_el2 before loading it
Alexandru Elisei [Wed, 7 Apr 2021 14:48:57 +0000 (15:48 +0100)]
KVM: arm64: Initialize VCPU mdcr_el2 before loading it

[ Upstream commit 263d6287da1433aba11c5b4046388f2cdf49675c ]

When a VCPU is created, the kvm_vcpu struct is initialized to zero in
kvm_vm_ioctl_create_vcpu(). On VHE systems, the first time
vcpu.arch.mdcr_el2 is loaded on hardware is in vcpu_load(), before it is
set to a sensible value in kvm_arm_setup_debug() later in the run loop. The
result is that KVM executes for a short time with MDCR_EL2 set to zero.

This has several unintended consequences:

* Setting MDCR_EL2.HPMN to 0 is constrained unpredictable according to ARM
  DDI 0487G.a, page D13-3820. The behavior specified by the architecture
  in this case is for the PE to behave as if MDCR_EL2.HPMN is set to a
  value less than or equal to PMCR_EL0.N, which means that an unknown
  number of counters are now disabled by MDCR_EL2.HPME, which is zero.

* The host configuration for the other debug features controlled by
  MDCR_EL2 is temporarily lost. This has been harmless so far, as Linux
  doesn't use the other fields, but that might change in the future.

Let's avoid both issues by initializing the VCPU's mdcr_el2 field in
kvm_vcpu_vcpu_first_run_init(), thus making sure that the MDCR_EL2 register
has a consistent value after each vcpu_load().

Fixes: d5a21bcc2995 ("KVM: arm64: Move common VHE/non-VHE trap config in separate functions")
Signed-off-by: Alexandru Elisei <alexandru.elisei@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210407144857.199746-3-alexandru.elisei@arm.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoHID: lenovo: Map mic-mute button to KEY_F20 instead of KEY_MICMUTE
Hans de Goede [Sun, 4 Apr 2021 08:04:29 +0000 (10:04 +0200)]
HID: lenovo: Map mic-mute button to KEY_F20 instead of KEY_MICMUTE

[ Upstream commit 617103246cfd19af837e4cb614ba9f877c4f7779 ]

Mapping the mic-mute button to KEY_MICMUTE is technically correct but
KEY_MICMUTE translates to a scancode of 256 (248 + 8) under X,
which does not fit in 8 bits, so it does not work.

Because of this userspace is expecting KEY_F20 instead,
theoretically KEY_MICMUTE should work under Wayland but even
there it does not work, because the desktop-environment is
listening only for KEY_F20 and not for KEY_MICMUTE.

Fixes: bc04b37ea0ec ("HID: lenovo: Add ThinkPad 10 Ultrabook Keyboard support")
Reviewed-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoHID: lenovo: Check hid_get_drvdata() returns non NULL in lenovo_event()
Hans de Goede [Sun, 4 Apr 2021 08:04:26 +0000 (10:04 +0200)]
HID: lenovo: Check hid_get_drvdata() returns non NULL in lenovo_event()

[ Upstream commit 34348a8661e3cd67dcf6938f08c8bb77522301f7 ]

The HID lenovo probe function only attaches drvdata to one of the
USB interfaces, but lenovo_event() will get called for all USB interfaces
to which hid-lenovo is bound.

This allows a malicious device to fake being a device handled by
hid-lenovo, which generates events for which lenovo_event() has
special handling (and thus dereferences hid_get_drvdata()) on another
interface triggering a NULL pointer exception.

Add a check for hid_get_drvdata() returning NULL, avoiding this
possible NULL pointer exception.

Fixes: bc04b37ea0ec ("HID: lenovo: Add ThinkPad 10 Ultrabook Keyboard support")
Reviewed-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoHID: lenovo: Fix lenovo_led_set_tp10ubkbd() error handling
Hans de Goede [Sun, 4 Apr 2021 08:04:25 +0000 (10:04 +0200)]
HID: lenovo: Fix lenovo_led_set_tp10ubkbd() error handling

[ Upstream commit 658d04e6eb6be1601ae95d7bee92bbf4096cdc1e ]

Fix the following issues with lenovo_led_set_tp10ubkbd() error handling:

1. On success hid_hw_raw_request() returns the number of bytes sent.
   So we should check for (ret != 3) rather then for (ret != 0).

2. Actually propagate errors to the caller.

3. Since the LEDs are part of an USB keyboard-dock the mute LEDs can go
   away at any time. Don't log an error when ret == -ENODEV and set the
   LED_HW_PLUGGABLE flag to avoid errors getting logged when the USB gets
   disconnected.

Fixes: bc04b37ea0ec ("HID: lenovo: Add ThinkPad 10 Ultrabook Keyboard support")
Reviewed-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoHID: lenovo: Use brightness_set_blocking callback for setting LEDs brightness
Hans de Goede [Sun, 4 Apr 2021 08:04:24 +0000 (10:04 +0200)]
HID: lenovo: Use brightness_set_blocking callback for setting LEDs brightness

[ Upstream commit bbf62645255f120bc2e7488c237e3f04da42ec70 ]

The lenovo_led_brightness_set function may sleep, so we should have the
the led_class_dev's brightness_set_blocking callback point to it, rather
then the regular brightness_set callback.

When toggled through sysfs this is not a problem, but the brightness_set
callback may be called from atomic context when using LED-triggers.

Fixes: bc04b37ea0ec ("HID: lenovo: Add ThinkPad 10 Ultrabook Keyboard support")
Reviewed-by: Marek Behún <kabel@kernel.org>
Acked-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoALSA: usb-audio: Add error checks for usb_driver_claim_interface() calls
Takashi Iwai [Tue, 6 Apr 2021 11:35:34 +0000 (13:35 +0200)]
ALSA: usb-audio: Add error checks for usb_driver_claim_interface() calls

[ Upstream commit 5fb45414ae03421255593fd5556aa2d1d82303aa ]

There are a few calls of usb_driver_claim_interface() but all of those
miss the proper error checks, as reported by Coverity.  This patch
adds those missing checks.

Along with it, replace the magic pointer with -1 with a constant
USB_AUDIO_IFACE_UNUSED for better readability.

Reported-by: coverity-bot <keescook+coverity-bot@chromium.org>
Addresses-Coverity-ID: 1475943 ("Error handling issues")
Addresses-Coverity-ID: 1475944 ("Error handling issues")
Addresses-Coverity-ID: 1475945 ("Error handling issues")
Fixes: b1ce7ba619d9 ("ALSA: usb-audio: claim autodetected PCM interfaces all at once")
Fixes: e5779998bf8b ("ALSA: usb-audio: refactor code")
Link: https://lore.kernel.org/r/202104051059.FB7F3016@keescook
Link: https://lore.kernel.org/r/20210406113534.30455-1-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Invalidate PASID cache when root/context entry changed
Lu Baolu [Sat, 20 Mar 2021 02:54:13 +0000 (10:54 +0800)]
iommu/vt-d: Invalidate PASID cache when root/context entry changed

[ Upstream commit c0474a606ecb9326227b4d68059942f9db88a897 ]

When the Intel IOMMU is operating in the scalable mode, some information
from the root and context table may be used to tag entries in the PASID
cache. Software should invalidate the PASID-cache when changing root or
context table entries.

Suggested-by: Ashok Raj <ashok.raj@intel.com>
Fixes: 7373a8cc38197 ("iommu/vt-d: Setup context and enable RID2PASID support")
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210320025415.641201-4-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Remove WO permissions on second-level paging entries
Lu Baolu [Sat, 20 Mar 2021 02:54:12 +0000 (10:54 +0800)]
iommu/vt-d: Remove WO permissions on second-level paging entries

[ Upstream commit eea53c5816889ee8b64544fa2e9311a81184ff9c ]

When the first level page table is used for IOVA translation, it only
supports Read-Only and Read-Write permissions. The Write-Only permission
is not supported as the PRESENT bit (implying Read permission) should
always set. When using second level, we still give separate permissions
that allows WriteOnly which seems inconsistent and awkward. We want to
have consistent behavior. After moving to 1st level, we don't want things
to work sometimes, and break if we use 2nd level for the same mappings.
Hence remove this configuration.

Suggested-by: Ashok Raj <ashok.raj@intel.com>
Fixes: b802d070a52a1 ("iommu/vt-d: Use iova over first level")
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210320025415.641201-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Preset Access/Dirty bits for IOVA over FL
Lu Baolu [Fri, 15 Jan 2021 00:42:02 +0000 (08:42 +0800)]
iommu/vt-d: Preset Access/Dirty bits for IOVA over FL

[ Upstream commit a8ce9ebbecdfda3322bbcece6b3b25888217f8e3 ]

The Access/Dirty bits in the first level page table entry will be set
whenever a page table entry was used for address translation or write
permission was successfully translated. This is always true when using
the first-level page table for kernel IOVA. Instead of wasting hardware
cycles to update the certain bits, it's better to set them up at the
beginning.

Suggested-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210115004202.953965-1-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Report the right page fault address
Lu Baolu [Sat, 20 Mar 2021 02:54:11 +0000 (10:54 +0800)]
iommu/vt-d: Report the right page fault address

[ Upstream commit 03d205094af45bca4f8e0498c461a893aa3ec6d9 ]

The Address field of the Page Request Descriptor only keeps bit [63:12]
of the offending address. Convert it to a full address before reporting
it to device drivers.

Fixes: eb8d93ea3c1d3 ("iommu/vt-d: Report page request faults for guest SVA")
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210320025415.641201-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Report right snoop capability when using FL for IOVA
Lu Baolu [Tue, 30 Mar 2021 02:11:45 +0000 (10:11 +0800)]
iommu/vt-d: Report right snoop capability when using FL for IOVA

[ Upstream commit 6c00612d0cba10f7d0917cf1f73c945003ed4cd7 ]

The Intel VT-d driver checks wrong register to report snoop capablility
when using first level page table for GPA to HPA translation. This might
lead the IOMMU driver to say that it supports snooping control, but in
reality, it does not. Fix this by always setting PASID-table-entry.PGSNP
whenever a pasid entry is setting up for GPA to HPA translation so that
the IOMMU driver could report snoop capability as long as it runs in the
scalable mode.

Fixes: b802d070a52a1 ("iommu/vt-d: Use iova over first level")
Suggested-by: Rajesh Sankaran <rajesh.sankaran@intel.com>
Suggested-by: Kevin Tian <kevin.tian@intel.com>
Suggested-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210330021145.13824-1-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu: Fix a boundary issue to avoid performance drop
Xiang Chen [Thu, 25 Mar 2021 03:38:24 +0000 (11:38 +0800)]
iommu: Fix a boundary issue to avoid performance drop

[ Upstream commit 3431c3f660a39f6ced954548a59dba6541ce3eb1 ]

After the change of patch ("iommu: Switch gather->end to the
inclusive end"), the performace drops from 1600+K IOPS to 1200K in our
kunpeng ARM64 platform.
We find that the range [start1, end1) actually is joint from the range
[end1, end2), but it is considered as disjoint after the change,
so it needs more times of TLB sync, and spends more time on it.
So fix the boundary issue to avoid performance drop.

Fixes: 862c3715de8f ("iommu: Switch gather->end to the inclusive end")
Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com>
Acked-by: Will Deacon <will@kernel.org>
Link: https://lore.kernel.org/r/1616643504-120688-1-git-send-email-chenxiang66@hisilicon.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Don't set then clear private data in prq_event_thread()
Lu Baolu [Sat, 20 Mar 2021 02:41:56 +0000 (10:41 +0800)]
iommu/vt-d: Don't set then clear private data in prq_event_thread()

[ Upstream commit 1d421058c815d54113d9afdf6db3f995c788cf0d ]

The VT-d specification (section 7.6) requires that the value in the
Private Data field of a Page Group Response Descriptor must match
the value in the Private Data field of the respective Page Request
Descriptor.

The private data field of a page group response descriptor is set then
immediately cleared in prq_event_thread(). This breaks the rule defined
by the VT-d specification. Fix it by moving clearing code up.

Fixes: 5b438f4ba315d ("iommu/vt-d: Support page request in scalable mode")
Cc: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Liu Yi L <yi.l.liu@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210320024156.640798-1-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomips: bmips: fix syscon-reboot nodes
Álvaro Fernández Rojas [Sun, 14 Mar 2021 16:43:46 +0000 (17:43 +0100)]
mips: bmips: fix syscon-reboot nodes

[ Upstream commit cde58b861a1d365568588adda59d42351c0c4ad3 ]

Commit a23c4134955e added the clock controller nodes, incorrectly changing the
syscon-reboot nodes addresses.

Fixes: a23c4134955e ("MIPS: BMIPS: add clock controller nodes")
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonet: hns3: Limiting the scope of vector_ring_chain variable
Salil Mehta [Mon, 5 Apr 2021 17:28:25 +0000 (18:28 +0100)]
net: hns3: Limiting the scope of vector_ring_chain variable

[ Upstream commit d392ecd1bc29ae15b0e284d5f732c2d36f244271 ]

Limiting the scope of the variable vector_ring_chain to the block where it
is used.

Fixes: 424eb834a9be ("net: hns3: Unified HNS3 {VF|PF} Ethernet Driver for hip08 SoC")
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonfc: pn533: prevent potential memory corruption
Dan Carpenter [Fri, 2 Apr 2021 11:44:42 +0000 (14:44 +0300)]
nfc: pn533: prevent potential memory corruption

[ Upstream commit ca4d4c34ae9aa5c3c0da76662c5e549d2fc0cc86 ]

If the "type_a->nfcid_len" is too large then it would lead to memory
corruption in pn533_target_found_type_a() when we do:

memcpy(nfc_tgt->nfcid1, tgt_type_a->nfcid_data, nfc_tgt->nfcid1_len);

Fixes: c3b1e1e8a76f ("NFC: Export NFCID1 from pn533")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoRDMA/core: Fix corrupted SL on passive side
Håkon Bugge [Mon, 22 Mar 2021 13:35:32 +0000 (14:35 +0100)]
RDMA/core: Fix corrupted SL on passive side

[ Upstream commit 194f64a3cad3ab9e381e996a13089de3215d1887 ]

On RoCE systems, a CM REQ contains a Primary Hop Limit > 1 and Primary
Subnet Local is zero.

In cm_req_handler(), the cm_process_routed_req() function is called. Since
the Primary Subnet Local value is zero in the request, and since this is
RoCE (Primary Local LID is permissive), the following statement will be
executed:

      IBA_SET(CM_REQ_PRIMARY_SL, req_msg, wc->sl);

This corrupts SL in req_msg if it was different from zero. In other words,
a request to setup a connection using an SL != zero, will not be honored,
and a connection using SL zero will be created instead.

Fixed by not calling cm_process_routed_req() on RoCE systems, the
cm_process_route_req() is only for IB anyhow.

Fixes: 3971c9f6dbf2 ("IB/cm: Add interim support for routed paths")
Link: https://lore.kernel.org/r/1616420132-31005-1-git-send-email-haakon.bugge@oracle.com
Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agobug: Remove redundant condition check in report_bug
Andrew Scull [Thu, 18 Mar 2021 14:33:07 +0000 (14:33 +0000)]
bug: Remove redundant condition check in report_bug

[ Upstream commit 3ad1a6cb0abc63d036fc866bd7c2c5983516dec5 ]

report_bug() will return early if it cannot find a bug corresponding to
the provided address. The subsequent test for the bug will always be
true so remove it.

Fixes: 1b4cfe3c0a30d ("lib/bug.c: exclude non-BUG/WARN exceptions from report_bug()")
Signed-off-by: Andrew Scull <ascull@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: "Steven Rostedt (VMware)" <rostedt@goodmis.org>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20210318143311.839894-2-ascull@google.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonet/tipc: fix missing destroy_workqueue() on error in tipc_crypto_start()
Yang Yingliang [Wed, 31 Mar 2021 08:36:02 +0000 (16:36 +0800)]
net/tipc: fix missing destroy_workqueue() on error in tipc_crypto_start()

[ Upstream commit ac1db7acea67777be1ba86e36e058c479eab6508 ]

Add the missing destroy_workqueue() before return from
tipc_crypto_start() in the error handling case.

Fixes: 1ef6f7c9390f ("tipc: add automatic session key exchange")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc/pseries: Only register vio drivers if vio bus exists
Michael Ellerman [Tue, 16 Mar 2021 01:09:38 +0000 (12:09 +1100)]
powerpc/pseries: Only register vio drivers if vio bus exists

[ Upstream commit 11d92156f7a862091009d7655d19c1e7de37fc7a ]

The vio bus is a fake bus, which we use on pseries LPARs (guests) to
discover devices provided by the hypervisor. There's no need or sense
in creating the vio bus on bare metal systems.

Which is why commit 4336b9337824 ("powerpc/pseries: Make vio and
ibmebus initcalls pseries specific") made the initialisation of the
vio bus only happen in LPARs.

However as a result of that commit we now see errors at boot on bare
metal systems:

  Driver 'hvc_console' was unable to register with bus_type 'vio' because the bus was not initialized.
  Driver 'tpm_ibmvtpm' was unable to register with bus_type 'vio' because the bus was not initialized.

This happens because those drivers are built-in, and are calling
vio_register_driver(). It in turn calls driver_register() with a
reference to vio_bus_type, but we haven't registered vio_bus_type with
the driver core.

Fix it by also guarding vio_register_driver() with a check to see if
we are on pseries.

Fixes: 4336b9337824 ("powerpc/pseries: Make vio and ibmebus initcalls pseries specific")
Reported-by: Paul Menzel <pmenzel@molgen.mpg.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-by: Tyrel Datwyler <tyreld@linux.ibm.com>
Link: https://lore.kernel.org/r/20210316010938.525657-1-mpe@ellerman.id.au
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoudp: never accept GSO_FRAGLIST packets
Paolo Abeni [Tue, 30 Mar 2021 10:28:52 +0000 (12:28 +0200)]
udp: never accept GSO_FRAGLIST packets

[ Upstream commit 78352f73dc5047f3f744764cc45912498c52f3c9 ]

Currently the UDP protocol delivers GSO_FRAGLIST packets to
the sockets without the expected segmentation.

This change addresses the issue introducing and maintaining
a couple of new fields to explicitly accept SKB_GSO_UDP_L4
or GSO_FRAGLIST packets. Additionally updates  udp_unexpected_gso()
accordingly.

UDP sockets enabling UDP_GRO stil keep accept_udp_fraglist
zeroed.

v1 -> v2:
 - use 2 bits instead of a whole GSO bitmask (Willem)

Fixes: 9fd1ff5d2ac7 ("udp: Support UDP fraglist GRO/GSO.")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonet: phy: lan87xx: fix access to wrong register of LAN87xx
Andre Edich [Mon, 29 Mar 2021 09:45:36 +0000 (11:45 +0200)]
net: phy: lan87xx: fix access to wrong register of LAN87xx

[ Upstream commit fdb5cc6ab3b6a1c0122d3644a63ef9dc7a610d35 ]

The function lan87xx_config_aneg_ext was introduced to configure
LAN95xxA but as well writes to undocumented register of LAN87xx.
This fix prevents that access.

The function lan87xx_config_aneg_ext gets more suitable for the new
behavior name.

Reported-by: Måns Rullgård <mans@mansr.com>
Fixes: 05b35e7eb9a1 ("smsc95xx: add phylib support")
Signed-off-by: Andre Edich <andre.edich@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoALSA: core: remove redundant spin_lock pair in snd_card_disconnect
Jia Zhou [Tue, 30 Mar 2021 11:19:23 +0000 (13:19 +0200)]
ALSA: core: remove redundant spin_lock pair in snd_card_disconnect

[ Upstream commit abc21649b3e5c34b143bf86f0c78e33d5815e250 ]

modification in commit 2a3f7221acdd ("ALSA: core: Fix card races between
register and disconnect") resulting in this problem.

Fixes: 2a3f7221acdd ("ALSA: core: Fix card races between register and disconnect")
Signed-off-by: Jia Zhou <zhou.jia2@zte.com.cn>
Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
Link: https://lore.kernel.org/r/1616989007-34429-1-git-send-email-wang.yi59@zte.com.cn
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agogpio: guard gpiochip_irqchip_add_domain() with GPIOLIB_IRQCHIP
Álvaro Fernández Rojas [Wed, 24 Mar 2021 08:19:02 +0000 (09:19 +0100)]
gpio: guard gpiochip_irqchip_add_domain() with GPIOLIB_IRQCHIP

[ Upstream commit 9c7d24693d864f90b27aad5d15fbfe226c02898b ]

The current code doesn't check if GPIOLIB_IRQCHIP is enabled, which results in
a compilation error when trying to build gpio-regmap if CONFIG_GPIOLIB_IRQCHIP
isn't enabled.

Fixes: 6a45b0e2589f ("gpiolib: Introduce gpiochip_irqchip_add_domain()")
Suggested-by: Michael Walle <michael@walle.cc>
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Michael Walle <michael@walle.cc>
Acked-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Link: https://lore.kernel.org/r/20210324081923.20379-2-noltari@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoMIPS/bpf: Enable bpf_probe_read{, str}() on MIPS again
Tiezhu Yang [Thu, 25 Mar 2021 12:50:01 +0000 (20:50 +0800)]
MIPS/bpf: Enable bpf_probe_read{, str}() on MIPS again

[ Upstream commit 66633abd0642f1e89d26e15f36fb13d3a1c535ff ]

After commit 0ebeea8ca8a4 ("bpf: Restrict bpf_probe_read{, str}() only to
archs where they work"), bpf_probe_read{, str}() functions were no longer
available on MIPS, so there exist some errors when running bpf program:

root@linux:/home/loongson/bcc# python examples/tracing/task_switch.py
bpf: Failed to load program: Invalid argument
[...]
11: (85) call bpf_probe_read#4
unknown func bpf_probe_read#4
[...]
Exception: Failed to load BPF program count_sched: Invalid argument

ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE should be restricted to archs
with non-overlapping address ranges, but they can overlap in EVA mode
on MIPS, so select ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE if !EVA in
arch/mips/Kconfig, otherwise the bpf old helper bpf_probe_read() will
not be available.

This is similar with the commit d195b1d1d119 ("powerpc/bpf: Enable
bpf_probe_read{, str}() on powerpc again").

Fixes: 0ebeea8ca8a4 ("bpf: Restrict bpf_probe_read{, str}() only to archs where they work")
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc: Fix HAVE_HARDLOCKUP_DETECTOR_ARCH build configuration
Chen Huang [Sat, 27 Mar 2021 09:49:00 +0000 (09:49 +0000)]
powerpc: Fix HAVE_HARDLOCKUP_DETECTOR_ARCH build configuration

[ Upstream commit 4fe529449d85e78972fa327999961ecc83a0b6db ]

When compiling the powerpc with the SMP disabled, it shows the issue:

arch/powerpc/kernel/watchdog.c: In function ‘watchdog_smp_panic’:
arch/powerpc/kernel/watchdog.c:177:4: error: implicit declaration of function ‘smp_send_nmi_ipi’; did you mean ‘smp_send_stop’? [-Werror=implicit-function-declaration]
  177 |    smp_send_nmi_ipi(c, wd_lockup_ipi, 1000000);
      |    ^~~~~~~~~~~~~~~~
      |    smp_send_stop
cc1: all warnings being treated as errors
make[2]: *** [scripts/Makefile.build:273: arch/powerpc/kernel/watchdog.o] Error 1
make[1]: *** [scripts/Makefile.build:534: arch/powerpc/kernel] Error 2
make: *** [Makefile:1980: arch/powerpc] Error 2
make: *** Waiting for unfinished jobs....

We found that powerpc used ipi to implement hardlockup watchdog, so the
HAVE_HARDLOCKUP_DETECTOR_ARCH should depend on the SMP.

Fixes: 2104180a5369 ("powerpc/64s: implement arch-specific hardlockup watchdog")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Chen Huang <chenhuang5@huawei.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210327094900.938555-1-chenhuang5@huawei.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoIB/isert: Fix a use after free in isert_connect_request
Lv Yunlong [Mon, 22 Mar 2021 16:13:25 +0000 (09:13 -0700)]
IB/isert: Fix a use after free in isert_connect_request

[ Upstream commit adb76a520d068a54ee5ca82e756cf8e5a47363a4 ]

The device is got by isert_device_get() with refcount is 1, and is
assigned to isert_conn by
  isert_conn->device = device.

When isert_create_qp() failed, device will be freed with
isert_device_put().

Later, the device is used in isert_free_login_buf(isert_conn) by the
isert_conn->device->ib_device statement.

Free the device in the correct order.

Fixes: ae9ea9ed38c9 ("iser-target: Split some logic in isert_connect_request to routines")
Link: https://lore.kernel.org/r/20210322161325.7491-1-lyl2019@mail.ustc.edu.cn
Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Max Gurtovoy <mgurtovoy@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoRDMA/mlx5: Fix drop packet rule in egress table
Maor Gottlieb [Thu, 18 Mar 2021 13:51:23 +0000 (15:51 +0200)]
RDMA/mlx5: Fix drop packet rule in egress table

[ Upstream commit c73700806d4e430d182c2be069d230076818a99a ]

Initial drop action support missed that drop action can be added to egress
flow tables as well. Add the missing support.

This requires making sure that dest_type isn't set to PORT which in turn
exposes a possibility of passing dst while indicating number of dsts as
zero. Explicitly check for number of dsts and pass the appropriate
pointer.

Fixes: f29de9eee782 ("RDMA/mlx5: Add support for drop action in DV steering")
Link: https://lore.kernel.org/r/20210318135123.680759-1-leon@kernel.org
Reviewed-by: Mark Bloch <markb@nvidia.com>
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/arm-smmu-v3: add bit field SFM into GERROR_ERR_MASK
Zhen Lei [Wed, 24 Mar 2021 08:16:03 +0000 (16:16 +0800)]
iommu/arm-smmu-v3: add bit field SFM into GERROR_ERR_MASK

[ Upstream commit 655c447c97d7fe462e6cd9e15809037be028bc70 ]

In arm_smmu_gerror_handler(), the value of the SMMU_GERROR register is
filtered by GERROR_ERR_MASK. However, the GERROR_ERR_MASK does not contain
the SFM bit. As a result, the subsequent error processing is not performed
when only the SFM error occurs.

Fixes: 48ec83bcbcf5 ("iommu/arm-smmu: Add initial driver support for ARM SMMUv3 devices")
Reported-by: Rui Zhu <zhurui3@huawei.com>
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Link: https://lore.kernel.org/r/20210324081603.1074-1-thunder.leizhen@huawei.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoASoC: wm8960: Remove bitclk relax condition in wm8960_configure_sysclk
Shengjiu Wang [Wed, 3 Mar 2021 03:07:42 +0000 (11:07 +0800)]
ASoC: wm8960: Remove bitclk relax condition in wm8960_configure_sysclk

[ Upstream commit 99067c07e8d877035f6249d194a317c78b7d052d ]

The call sequence in wm8960_configure_clocking is

   ret = wm8960_configure_sysclk();
   if (ret >= 0)
        goto configure_clock;

   ....

   ret = wm8960_configure_pll();

configure_clock:
   ...

wm8960_configure_sysclk is called before wm8960_configure_pll, as
there is bitclk relax on both functions, so wm8960_configure_sysclk
always return success, then wm8960_configure_pll() never be called.

With this case:
aplay -Dhw:0,0 -d 5 -r 48000 -f S24_LE -c 2 audio48k24b2c.wav
the required bitclk is 48000 * 24 * 2 = 2304000, bitclk got from
wm8960_configure_sysclk is 3072000, but if go to wm8960_configure_pll.
it can get correct bitclk 2304000.

So bitclk relax condition should be removed in wm8960_configure_sysclk,
then wm8960_configure_pll can be called, and there is also bitclk relax
function in wm8960_configure_pll.

Fixes: 3c01b9ee2ab9 ("ASoC: codec: wm8960: Relax bit clock computation")
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/1614740862-30196-1-git-send-email-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoMIPS: loongson64: fix bug when PAGE_SIZE > 16KB
Huang Pei [Wed, 24 Mar 2021 03:24:51 +0000 (11:24 +0800)]
MIPS: loongson64: fix bug when PAGE_SIZE > 16KB

[ Upstream commit 509d36a941a3466b78d4377913623d210b162458 ]

When page size larger than 16KB, arguments "vaddr + size(16KB)" in
"ioremap_page_range(vaddr, vaddr + size,...)" called by
"add_legacy_isa_io" is not page-aligned.

As loongson64 needs at least page size 16KB to get rid of cache alias,
and "vaddr" is 64KB-aligned, and 64KB is largest page size supported,
rounding "size" up to PAGE_SIZE is enough for all page size supported.

Fixes: 6d0068ad15e4 ("MIPS: Loongson64: Process ISA Node in DeviceTree")
Signed-off-by: Huang Pei <huangpei@loongson.cn>
Acked-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopinctrl: pinctrl-single: fix pcs_pin_dbg_show() when bits_per_mux is not zero
Hanna Hawa [Fri, 19 Mar 2021 15:21:33 +0000 (17:21 +0200)]
pinctrl: pinctrl-single: fix pcs_pin_dbg_show() when bits_per_mux is not zero

[ Upstream commit bd85125ea88513f637a62a72e8949c579c5c0a87 ]

A System Error (SError, followed by kernel panic) was detected when
trying to print the supported pins in a pinctrl device which supports
multiple pins per register. This change fixes the pcs_pin_dbg_show() in
pinctrl-single driver when bits_per_mux is not zero. In addition move
offset calculation and pin offset in register to common function.

Fixes: 4e7e8017a80e ("pinctrl: pinctrl-single: enhance to configure multiple pins of different modules")
Signed-off-by: Hanna Hawa <hhhawa@amazon.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Reviewed-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Drew Fustini <drew@beagleboard.org>
Link: https://lore.kernel.org/r/20210319152133.28705-4-hhhawa@amazon.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopinctrl: pinctrl-single: remove unused parameter
Hanna Hawa [Fri, 19 Mar 2021 15:21:32 +0000 (17:21 +0200)]
pinctrl: pinctrl-single: remove unused parameter

[ Upstream commit 8fa2ea202b13b6da81e26c399ff1d87488398453 ]

Remove unused parameter 'pin_pos' from pcs_add_pin().

Signed-off-by: Hanna Hawa <hhhawa@amazon.com>
Reviewed-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Drew Fustini <drew@beagleboard.org>
Link: https://lore.kernel.org/r/20210319152133.28705-3-hhhawa@amazon.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoinet: use bigger hash table for IP ID generation
Eric Dumazet [Wed, 24 Mar 2021 21:53:37 +0000 (14:53 -0700)]
inet: use bigger hash table for IP ID generation

[ Upstream commit aa6dd211e4b1dde9d5dc25d699d35f789ae7eeba ]

In commit 73f156a6e8c1 ("inetpeer: get rid of ip_id_count")
I used a very small hash table that could be abused
by patient attackers to reveal sensitive information.

Switch to a dynamic sizing, depending on RAM size.

Typical big hosts will now use 128x more storage (2 MB)
to get a similar increase in security and reduction
of hash collisions.

As a bonus, use of alloc_large_system_hash() spreads
allocated memory among all NUMA nodes.

Fixes: 73f156a6e8c1 ("inetpeer: get rid of ip_id_count")
Reported-by: Amit Klein <aksecurity@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willy Tarreau <w@1wt.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoima: Fix the error code for restoring the PCR value
Li Huafei [Wed, 3 Mar 2021 03:28:24 +0000 (11:28 +0800)]
ima: Fix the error code for restoring the PCR value

[ Upstream commit 7990ccafaa37dc6d8bb095d4d7cd997e8903fd10 ]

In ima_restore_measurement_list(), hdr[HDR_PCR].data is pointing to a
buffer of type u8, which contains the dumped 32-bit pcr value.
Currently, only the least significant byte is used to restore the pcr
value. We should convert hdr[HDR_PCR].data to a pointer of type u32
before fetching the value to restore the correct pcr value.

Fixes: 47fdee60b47f ("ima: use ima_parse_buf() to parse measurements headers")
Signed-off-by: Li Huafei <lihuafei1@huawei.com>
Reviewed-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoMIPS: fix local_irq_{disable,enable} in asmmacro.h
Huang Pei [Tue, 23 Mar 2021 02:34:02 +0000 (10:34 +0800)]
MIPS: fix local_irq_{disable,enable} in asmmacro.h

[ Upstream commit 05c4e2721d7af0df7bc1378a23712a0fd16947b5 ]

commit ba9196d2e005 ("MIPS: Make DIEI support as a config option")
use CPU_HAS_DIEI to indicate whether di/ei is implemented correctly,
without this patch, "local_irq_disable" from entry.S in 3A1000
(with buggy di/ei) lose protection of commit e97c5b609880 ("MIPS:
Make irqflags.h functions preempt-safe for non-mipsr2 cpus")

Fixes: ba9196d2e005 ("MIPS: Make DIEI support as a config option")
Signed-off-by: Huang Pei <huangpei@loongson.cn>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc/prom: Mark identical_pvr_fixup as __init
Nathan Chancellor [Tue, 2 Mar 2021 20:08:29 +0000 (13:08 -0700)]
powerpc/prom: Mark identical_pvr_fixup as __init

[ Upstream commit 1ef1dd9c7ed27b080445e1576e8a05957e0e4dfc ]

If identical_pvr_fixup() is not inlined, there are two modpost warnings:

WARNING: modpost: vmlinux.o(.text+0x54e8): Section mismatch in reference
from the function identical_pvr_fixup() to the function
.init.text:of_get_flat_dt_prop()
The function identical_pvr_fixup() references
the function __init of_get_flat_dt_prop().
This is often because identical_pvr_fixup lacks a __init
annotation or the annotation of of_get_flat_dt_prop is wrong.

WARNING: modpost: vmlinux.o(.text+0x551c): Section mismatch in reference
from the function identical_pvr_fixup() to the function
.init.text:identify_cpu()
The function identical_pvr_fixup() references
the function __init identify_cpu().
This is often because identical_pvr_fixup lacks a __init
annotation or the annotation of identify_cpu is wrong.

identical_pvr_fixup() calls two functions marked as __init and is only
called by a function marked as __init so it should be marked as __init
as well. At the same time, remove the inline keywork as it is not
necessary to inline this function. The compiler is still free to do so
if it feels it is worthwhile since commit 889b3c1245de ("compiler:
remove CONFIG_OPTIMIZE_INLINING entirely").

Fixes: 14b3d926a22b ("[POWERPC] 4xx: update 440EP(x)/440GR(x) identical PVR issue workaround")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://github.com/ClangBuiltLinux/linux/issues/1316
Link: https://lore.kernel.org/r/20210302200829.2680663-1-nathan@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agopowerpc/fadump: Mark fadump_calculate_reserve_size as __init
Nathan Chancellor [Tue, 2 Mar 2021 19:50:14 +0000 (12:50 -0700)]
powerpc/fadump: Mark fadump_calculate_reserve_size as __init

[ Upstream commit fbced1546eaaab57a32e56c974ea8acf10c6abd8 ]

If fadump_calculate_reserve_size() is not inlined, there is a modpost
warning:

WARNING: modpost: vmlinux.o(.text+0x5196c): Section mismatch in
reference from the function fadump_calculate_reserve_size() to the
function .init.text:parse_crashkernel()
The function fadump_calculate_reserve_size() references
the function __init parse_crashkernel().
This is often because fadump_calculate_reserve_size lacks a __init
annotation or the annotation of parse_crashkernel is wrong.

fadump_calculate_reserve_size() calls parse_crashkernel(), which is
marked as __init and fadump_calculate_reserve_size() is called from
within fadump_reserve_mem(), which is also marked as __init.

Mark fadump_calculate_reserve_size() as __init to fix the section
mismatch. Additionally, remove the inline keyword as it is not necessary
to inline this function; the compiler is still free to do so if it feels
it is worthwhile since commit 889b3c1245de ("compiler: remove
CONFIG_OPTIMIZE_INLINING entirely").

Fixes: 11550dc0a00b ("powerpc/fadump: reuse crashkernel parameter for fadump memory reservation")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://github.com/ClangBuiltLinux/linux/issues/1300
Link: https://lore.kernel.org/r/20210302195013.2626335-1-nathan@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agolibbpf: Add explicit padding to btf_dump_emit_type_decl_opts
KP Singh [Fri, 19 Mar 2021 19:21:17 +0000 (19:21 +0000)]
libbpf: Add explicit padding to btf_dump_emit_type_decl_opts

[ Upstream commit ea24b19562fe5f72c78319dbb347b701818956d9 ]

Similar to
https://lore.kernel.org/bpf/20210313210920.1959628-2-andrii@kernel.org/

When DECLARE_LIBBPF_OPTS is used with inline field initialization, e.g:

  DECLARE_LIBBPF_OPTS(btf_dump_emit_type_decl_opts, opts,
    .field_name = var_ident,
    .indent_level = 2,
    .strip_mods = strip_mods,
  );

and compiled in debug mode, the compiler generates code which
leaves the padding uninitialized and triggers errors within libbpf APIs
which require strict zero initialization of OPTS structs.

Adding anonymous padding field fixes the issue.

Fixes: 9f81654eebe8 ("libbpf: Expose BTF-to-C type declaration emitting API")
Suggested-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: KP Singh <kpsingh@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210319192117.2310658-1-kpsingh@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoselftests/bpf: Re-generate vmlinux.h and BPF skeletons if bpftool changed
Andrii Nakryiko [Thu, 18 Mar 2021 19:40:34 +0000 (12:40 -0700)]
selftests/bpf: Re-generate vmlinux.h and BPF skeletons if bpftool changed

[ Upstream commit cab62c37be057379a2a17b1b2eacd9dcba1e14dc ]

Trigger vmlinux.h and BPF skeletons re-generation if detected that bpftool was
re-compiled. Otherwise full `make clean` is required to get updated skeletons,
if bpftool is modified.

Fixes: acbd06206bbb ("selftests/bpf: Add vmlinux.h selftest exercising tracing of syscalls")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210318194036.3521577-11-andrii@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu/vt-d: Reject unsupported page request modes
Jacob Pan [Tue, 2 Mar 2021 10:13:59 +0000 (02:13 -0800)]
iommu/vt-d: Reject unsupported page request modes

[ Upstream commit 78a523fe73b81b4447beb2d6c78c9fafae24eebb ]

When supervisor/privilige mode SVM is used, we bind init_mm.pgd with
a supervisor PASID. There should not be any page fault for init_mm.
Execution request with DMA read is also not supported.

This patch checks PRQ descriptor for both unsupported configurations,
reject them both with invalid responses.

Fixes: 1c4f88b7f1f92 ("iommu/vt-d: Shared virtual address in scalable mode")
Acked-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Link: https://lore.kernel.org/r/1614680040-1989-4-git-send-email-jacob.jun.pan@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoiommu: Check dev->iommu in iommu_dev_xxx functions
Shameer Kolothum [Wed, 3 Mar 2021 17:36:11 +0000 (17:36 +0000)]
iommu: Check dev->iommu in iommu_dev_xxx functions

[ Upstream commit b9abb19fa5fd2d8a4be61c6cd4b2a48aa1a17f9c ]

The device iommu probe/attach might have failed leaving dev->iommu
to NULL and device drivers may still invoke these functions resulting
in a crash in iommu vendor driver code.

Hence make sure we check that.

Fixes: a3a195929d40 ("iommu: Add APIs for multiple domains per device")
Signed-off-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/20210303173611.520-1-shameerali.kolothum.thodi@huawei.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agobpftool: Fix maybe-uninitialized warnings
Andrii Nakryiko [Sat, 13 Mar 2021 21:09:18 +0000 (13:09 -0800)]
bpftool: Fix maybe-uninitialized warnings

[ Upstream commit 4bbb3583687051ef99966ddaeb1730441b777d40 ]

Somehow when bpftool is compiled in -Og mode, compiler produces new warnings
about possibly uninitialized variables. Fix all the reported problems.

Fixes: 2119f2189df1 ("bpftool: add C output format option to btf dump subcommand")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210313210920.1959628-3-andrii@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agolibbpf: Add explicit padding to bpf_xdp_set_link_opts
Andrii Nakryiko [Sat, 13 Mar 2021 21:09:17 +0000 (13:09 -0800)]
libbpf: Add explicit padding to bpf_xdp_set_link_opts

[ Upstream commit dde7b3f5f2f458297aeccfd4783e53ab8ca046db ]

Adding such anonymous padding fixes the issue with uninitialized portions of
bpf_xdp_set_link_opts when using LIBBPF_DECLARE_OPTS macro with inline field
initialization:

DECLARE_LIBBPF_OPTS(bpf_xdp_set_link_opts, opts, .old_fd = -1);

When such code is compiled in debug mode, compiler is generating code that
leaves padding bytes uninitialized, which triggers error inside libbpf APIs
that do strict zero initialization checks for OPTS structs.

Adding anonymous padding field fixes the issue.

Fixes: bd5ca3ef93cd ("libbpf: Add function to set link XDP fd while specifying old program")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210313210920.1959628-2-andrii@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonet: lapbether: Prevent racing when checking whether the netif is running
Xie He [Thu, 11 Mar 2021 07:23:09 +0000 (23:23 -0800)]
net: lapbether: Prevent racing when checking whether the netif is running

[ Upstream commit 5acd0cfbfbb5a688da1bfb1a2152b0c855115a35 ]

There are two "netif_running" checks in this driver. One is in
"lapbeth_xmit" and the other is in "lapbeth_rcv". They serve to make
sure that the LAPB APIs called in these functions are called before
"lapb_unregister" is called by the "ndo_stop" function.

However, these "netif_running" checks are unreliable, because it's
possible that immediately after "netif_running" returns true, "ndo_stop"
is called (which causes "lapb_unregister" to be called).

This patch adds locking to make sure "lapbeth_xmit" and "lapbeth_rcv" can
reliably check and ensure the netif is running while doing their work.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Acked-by: Martin Schiller <ms@dev.tdt.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoBluetooth: avoid deadlock between hci_dev->lock and socket lock
Jiri Kosina [Tue, 16 Mar 2021 14:08:00 +0000 (15:08 +0100)]
Bluetooth: avoid deadlock between hci_dev->lock and socket lock

[ Upstream commit 17486960d79b900c45e0bb8fbcac0262848582ba ]

Commit eab2404ba798 ("Bluetooth: Add BT_PHY socket option") added a
dependency between socket lock and hci_dev->lock that could lead to
deadlock.

It turns out that hci_conn_get_phy() is not in any way relying on hdev
being immutable during the runtime of this function, neither does it even
look at any of the members of hdev, and as such there is no need to hold
that lock.

This fixes the lockdep splat below:

 ======================================================
 WARNING: possible circular locking dependency detected
 5.12.0-rc1-00026-g73d464503354 #10 Not tainted
 ------------------------------------------------------
 bluetoothd/1118 is trying to acquire lock:
 ffff8f078383c078 (&hdev->lock){+.+.}-{3:3}, at: hci_conn_get_phy+0x1c/0x150 [bluetooth]

 but task is already holding lock:
 ffff8f07e831d920 (sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP){+.+.}-{0:0}, at: l2cap_sock_getsockopt+0x8b/0x610

 which lock already depends on the new lock.

 the existing dependency chain (in reverse order) is:

 -> #3 (sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP){+.+.}-{0:0}:
        lock_sock_nested+0x72/0xa0
        l2cap_sock_ready_cb+0x18/0x70 [bluetooth]
        l2cap_config_rsp+0x27a/0x520 [bluetooth]
        l2cap_sig_channel+0x658/0x1330 [bluetooth]
        l2cap_recv_frame+0x1ba/0x310 [bluetooth]
        hci_rx_work+0x1cc/0x640 [bluetooth]
        process_one_work+0x244/0x5f0
        worker_thread+0x3c/0x380
        kthread+0x13e/0x160
        ret_from_fork+0x22/0x30

 -> #2 (&chan->lock#2/1){+.+.}-{3:3}:
        __mutex_lock+0xa3/0xa10
        l2cap_chan_connect+0x33a/0x940 [bluetooth]
        l2cap_sock_connect+0x141/0x2a0 [bluetooth]
        __sys_connect+0x9b/0xc0
        __x64_sys_connect+0x16/0x20
        do_syscall_64+0x33/0x80
        entry_SYSCALL_64_after_hwframe+0x44/0xae

 -> #1 (&conn->chan_lock){+.+.}-{3:3}:
        __mutex_lock+0xa3/0xa10
        l2cap_chan_connect+0x322/0x940 [bluetooth]
        l2cap_sock_connect+0x141/0x2a0 [bluetooth]
        __sys_connect+0x9b/0xc0
        __x64_sys_connect+0x16/0x20
        do_syscall_64+0x33/0x80
        entry_SYSCALL_64_after_hwframe+0x44/0xae

 -> #0 (&hdev->lock){+.+.}-{3:3}:
        __lock_acquire+0x147a/0x1a50
        lock_acquire+0x277/0x3d0
        __mutex_lock+0xa3/0xa10
        hci_conn_get_phy+0x1c/0x150 [bluetooth]
        l2cap_sock_getsockopt+0x5a9/0x610 [bluetooth]
        __sys_getsockopt+0xcc/0x200
        __x64_sys_getsockopt+0x20/0x30
        do_syscall_64+0x33/0x80
        entry_SYSCALL_64_after_hwframe+0x44/0xae

 other info that might help us debug this:

 Chain exists of:
   &hdev->lock --> &chan->lock#2/1 --> sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP

  Possible unsafe locking scenario:

        CPU0                    CPU1
        ----                    ----
   lock(sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP);
                                lock(&chan->lock#2/1);
                                lock(sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP);
   lock(&hdev->lock);

  *** DEADLOCK ***

 1 lock held by bluetoothd/1118:
  #0: ffff8f07e831d920 (sk_lock-AF_BLUETOOTH-BTPROTO_L2CAP){+.+.}-{0:0}, at: l2cap_sock_getsockopt+0x8b/0x610 [bluetooth]

 stack backtrace:
 CPU: 3 PID: 1118 Comm: bluetoothd Not tainted 5.12.0-rc1-00026-g73d464503354 #10
 Hardware name: LENOVO 20K5S22R00/20K5S22R00, BIOS R0IET38W (1.16 ) 05/31/2017
 Call Trace:
  dump_stack+0x7f/0xa1
  check_noncircular+0x105/0x120
  ? __lock_acquire+0x147a/0x1a50
  __lock_acquire+0x147a/0x1a50
  lock_acquire+0x277/0x3d0
  ? hci_conn_get_phy+0x1c/0x150 [bluetooth]
  ? __lock_acquire+0x2e1/0x1a50
  ? lock_is_held_type+0xb4/0x120
  ? hci_conn_get_phy+0x1c/0x150 [bluetooth]
  __mutex_lock+0xa3/0xa10
  ? hci_conn_get_phy+0x1c/0x150 [bluetooth]
  ? lock_acquire+0x277/0x3d0
  ? mark_held_locks+0x49/0x70
  ? mark_held_locks+0x49/0x70
  ? hci_conn_get_phy+0x1c/0x150 [bluetooth]
  hci_conn_get_phy+0x1c/0x150 [bluetooth]
  l2cap_sock_getsockopt+0x5a9/0x610 [bluetooth]
  __sys_getsockopt+0xcc/0x200
  __x64_sys_getsockopt+0x20/0x30
  do_syscall_64+0x33/0x80
  entry_SYSCALL_64_after_hwframe+0x44/0xae
 RIP: 0033:0x7fb73df33eee
 Code: 48 8b 0d 85 0f 0c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 49 89 ca b8 37 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 52 0f 0c 00 f7 d8 64 89 01 48
 RSP: 002b:00007fffcfbbbf08 EFLAGS: 00000203 ORIG_RAX: 0000000000000037
 RAX: ffffffffffffffda RBX: 0000000000000019 RCX: 00007fb73df33eee
 RDX: 000000000000000e RSI: 0000000000000112 RDI: 0000000000000018
 RBP: 0000000000000000 R08: 00007fffcfbbbf44 R09: 0000000000000000
 R10: 00007fffcfbbbf3c R11: 0000000000000203 R12: 0000000000000000
 R13: 0000000000000018 R14: 0000000000000000 R15: 0000556fcefc70d0

Fixes: eab2404ba798 ("Bluetooth: Add BT_PHY socket option")
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoKVM: x86/mmu: Retry page faults that hit an invalid memslot
Sean Christopherson [Thu, 25 Feb 2021 20:47:30 +0000 (12:47 -0800)]
KVM: x86/mmu: Retry page faults that hit an invalid memslot

[ Upstream commit e0c378684b6545ad2d4403bb701d0ac4932b4e95 ]

Retry page faults (re-enter the guest) that hit an invalid memslot
instead of treating the memslot as not existing, i.e. handling the
page fault as an MMIO access.  When deleting a memslot, SPTEs aren't
zapped and the TLBs aren't flushed until after the memslot has been
marked invalid.

Handling the invalid slot as MMIO means there's a small window where a
page fault could replace a valid SPTE with an MMIO SPTE.  The legacy
MMU handles such a scenario cleanly, but the TDP MMU assumes such
behavior is impossible (see the BUG() in __handle_changed_spte()).
There's really no good reason why the legacy MMU should allow such a
scenario, and closing this hole allows for additional cleanups.

Fixes: 2f2fad0897cb ("kvm: x86/mmu: Add functions to handle changed TDP SPTEs")
Cc: Ben Gardon <bgardon@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20210225204749.1512652-6-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agowilc1000: write value to WILC_INTR2_ENABLE register
Marcus Folkesson [Wed, 24 Feb 2021 16:37:06 +0000 (17:37 +0100)]
wilc1000: write value to WILC_INTR2_ENABLE register

[ Upstream commit e21b6e5a54628cd3935f200049d4430c25c54e03 ]

Write the value instead of reading it twice.

Fixes: c5c77ba18ea6 ("staging: wilc1000: Add SDIO/SPI 802.11 driver")
Signed-off-by: Marcus Folkesson <marcus.folkesson@gmail.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20210224163706.519658-1-marcus.folkesson@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoRDMA/mlx5: Fix mlx5 rates to IB rates map
Mark Zhang [Thu, 4 Mar 2021 12:45:17 +0000 (14:45 +0200)]
RDMA/mlx5: Fix mlx5 rates to IB rates map

[ Upstream commit 6fe6e568639859db960c8fcef19a2ece1c2d7eae ]

Correct the map between mlx5 rates and corresponding ib rates, as they
don't always have a fixed offset between them.

Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters")
Link: https://lore.kernel.org/r/20210304124517.1100608-4-leon@kernel.org
Signed-off-by: Mark Zhang <markzhang@nvidia.com>
Reviewed-by: Maor Gottlieb <maorg@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoASoC: Intel: Skylake: Compile when any configuration is selected
Cezary Rojewski [Mon, 25 Jan 2021 11:54:41 +0000 (12:54 +0100)]
ASoC: Intel: Skylake: Compile when any configuration is selected

[ Upstream commit 1b99d50b9709a2cddaba4a7faf1862b4f7bec865 ]

Skylake is dependent on SND_SOC_INTEL_SKYLAKE (aka "all SST platforms")
whereas selecting specific configuration such as KBL-only will not
cause driver code to compile. Switch to SND_SOC_INTEL_SKYLAKE_COMMON
dependency so selecting any configuration causes the driver to be built.

Reported-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Suggested-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Fixes: 35bc99aaa1a3 ("ASoC: Intel: Skylake: Add more platform granularity")
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20210125115441.10383-1-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoASoC: Intel: boards: sof-wm8804: add check for PLL setting
Colin Ian King [Fri, 26 Feb 2021 18:56:53 +0000 (18:56 +0000)]
ASoC: Intel: boards: sof-wm8804: add check for PLL setting

[ Upstream commit 1730ef62874dbdc53dc2abfa430f09f0b304bafc ]

Currently the return from snd_soc_dai_set_pll is not checking for
failure, this is the only driver in the kernel that ignores this,
so it probably should be added for sake of completeness.  Fix this
by adding an error return check.

Addresses-Coverity: ("Unchecked return value")
Fixes: f139546fb7d4 ("ASoC: Intel: boards: sof-wm8804: support for Hifiberry Digiplus boards")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210226185653.1071321-1-colin.king@canonical.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoperf symbols: Fix dso__fprintf_symbols_by_name() to return the number of printed...
Arnaldo Carvalho de Melo [Mon, 8 Mar 2021 14:17:51 +0000 (11:17 -0300)]
perf symbols: Fix dso__fprintf_symbols_by_name() to return the number of printed chars

[ Upstream commit 210e4c89ef61432040c6cd828fefa441f4887186 ]

The 'ret' variable was initialized to zero but then it was not updated
from the fprintf() return, fix it.

Reported-by: Yang Li <yang.lee@linux.alibaba.com>
cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
cc: Ingo Molnar <mingo@redhat.com>
cc: Jiri Olsa <jolsa@redhat.com>
cc: Mark Rutland <mark.rutland@arm.com>
cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Fixes: 90f18e63fbd00513 ("perf symbols: List symbols in a dso in ascending name order")
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoHID: plantronics: Workaround for double volume key presses
Maxim Mikityanskiy [Sun, 7 Feb 2021 14:47:40 +0000 (16:47 +0200)]
HID: plantronics: Workaround for double volume key presses

[ Upstream commit f567d6ef8606fb427636e824c867229ecb5aefab ]

Plantronics Blackwire 3220 Series (047f:c056) sends HID reports twice
for each volume key press. This patch adds a quirk to hid-plantronics
for this product ID, which will ignore the second volume key press if
it happens within 5 ms from the last one that was handled.

The patch was tested on the mentioned model only, it shouldn't affect
other models, however, this quirk might be needed for them too.
Auto-repeat (when a key is held pressed) is not affected, because the
rate is about 3 times per second, which is far less frequent than once
in 5 ms.

Fixes: 81bb773faed7 ("HID: plantronics: Update to map volume up/down controls")
Signed-off-by: Maxim Mikityanskiy <maxtram95@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoxsk: Respect device's headroom and tailroom on generic xmit path
Alexander Lobakin [Thu, 18 Feb 2021 20:50:31 +0000 (20:50 +0000)]
xsk: Respect device's headroom and tailroom on generic xmit path

[ Upstream commit 3914d88f7608e6c2e80e344474fa289370c32451 ]

xsk_generic_xmit() allocates a new skb and then queues it for
xmitting. The size of new skb's headroom is desc->len, so it comes
to the driver/device with no reserved headroom and/or tailroom.
Lots of drivers need some headroom (and sometimes tailroom) to
prepend (and/or append) some headers or data, e.g. CPU tags,
device-specific headers/descriptors (LSO, TLS etc.), and if case
of no available space skb_cow_head() will reallocate the skb.
Reallocations are unwanted on fast-path, especially when it comes
to XDP, so generic XSK xmit should reserve the spaces declared in
dev->needed_headroom and dev->needed tailroom to avoid them.

Note on max(NET_SKB_PAD, L1_CACHE_ALIGN(dev->needed_headroom)):

Usually, output functions reserve LL_RESERVED_SPACE(dev), which
consists of dev->hard_header_len + dev->needed_headroom, aligned
by 16.

However, on XSK xmit hard header is already here in the chunk, so
hard_header_len is not needed. But it'd still be better to align
data up to cacheline, while reserving no less than driver requests
for headroom. NET_SKB_PAD here is to double-insure there will be
no reallocations even when the driver advertises no needed_headroom,
but in fact need it (not so rare case).

Fixes: 35fcde7f8deb ("xsk: support for Tx")
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20210218204908.5455-5-alobakin@pm.me
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agodrivers/block/null_blk/main: Fix a double free in null_init.
Lv Yunlong [Mon, 26 Apr 2021 14:32:29 +0000 (07:32 -0700)]
drivers/block/null_blk/main: Fix a double free in null_init.

[ Upstream commit 72ce11ddfa4e9e1879103581a60b7e34547eaa0a ]

In null_init, null_add_dev(dev) is called.
In null_add_dev, it calls null_free_zoned_dev(dev) to free dev->zones
via kvfree(dev->zones) in out_cleanup_zone branch and returns err.
Then null_init accept the err code and then calls null_free_dev(dev).

But in null_free_dev(dev), dev->zones is freed again by
null_free_zoned_dev().

My patch set dev->zones to NULL in null_free_zoned_dev() after
kvfree(dev->zones) is called, to avoid the double free.

Fixes: 2984c8684f962 ("nullb: factor disk parameters")
Signed-off-by: Lv Yunlong <lyl2019@mail.ustc.edu.cn>
Link: https://lore.kernel.org/r/20210426143229.7374-1-lyl2019@mail.ustc.edu.cn
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agosched/debug: Fix cgroup_path[] serialization
Waiman Long [Thu, 15 Apr 2021 19:54:26 +0000 (15:54 -0400)]
sched/debug: Fix cgroup_path[] serialization

[ Upstream commit ad789f84c9a145f8a18744c0387cec22ec51651e ]

The handling of sysrq key can be activated by echoing the key to
/proc/sysrq-trigger or via the magic key sequence typed into a terminal
that is connected to the system in some way (serial, USB or other mean).
In the former case, the handling is done in a user context. In the
latter case, it is likely to be in an interrupt context.

Currently in print_cpu() of kernel/sched/debug.c, sched_debug_lock is
taken with interrupt disabled for the whole duration of the calls to
print_*_stats() and print_rq() which could last for the quite some time
if the information dump happens on the serial console.

If the system has many cpus and the sched_debug_lock is somehow busy
(e.g. parallel sysrq-t), the system may hit a hard lockup panic
depending on the actually serial console implementation of the
system.

The purpose of sched_debug_lock is to serialize the use of the global
cgroup_path[] buffer in print_cpu(). The rests of the printk calls don't
need serialization from sched_debug_lock.

Calling printk() with interrupt disabled can still be problematic if
multiple instances are running. Allocating a stack buffer of PATH_MAX
bytes is not feasible because of the limited size of the kernel stack.

The solution implemented in this patch is to allow only one caller at a
time to use the full size group_path[], while other simultaneous callers
will have to use shorter stack buffers with the possibility of path
name truncation. A "..." suffix will be printed if truncation may have
happened.  The cgroup path name is provided for informational purpose
only, so occasional path name truncation should not be a big problem.

Fixes: efe25c2c7b3a ("sched: Reinstate group names in /proc/sched_debug")
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Waiman Long <longman@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210415195426.6677-1-longman@redhat.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoio_uring: fix overflows checks in provide buffers
Pavel Begunkov [Thu, 15 Apr 2021 12:07:39 +0000 (13:07 +0100)]
io_uring: fix overflows checks in provide buffers

[ Upstream commit 38134ada0ceea3e848fe993263c0ff6207fd46e7 ]

Colin reported before possible overflow and sign extension problems in
io_provide_buffers_prep(). As Linus pointed out previous attempt did nothing
useful, see d81269fecb8ce ("io_uring: fix provide_buffers sign extension").

Do that with help of check_<op>_overflow helpers. And fix struct
io_provide_buf::len type, as it doesn't make much sense to keep it
signed.

Reported-by: Colin Ian King <colin.king@canonical.com>
Fixes: efe68c1ca8f49 ("io_uring: validate the full range of provided buffers for access")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/46538827e70fce5f6cdb50897cff4cacc490f380.1618488258.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoperf/amd/uncore: Fix sysfs type mismatch
Nathan Chancellor [Thu, 15 Apr 2021 00:11:12 +0000 (17:11 -0700)]
perf/amd/uncore: Fix sysfs type mismatch

[ Upstream commit 5deac80d4571dffb51f452f0027979d72259a1b9 ]

dev_attr_show() calls the __uncore_*_show() functions via an indirect
call but their type does not currently match the type of the show()
member in 'struct device_attribute', resulting in a Control Flow
Integrity violation.

$ cat /sys/devices/amd_l3/format/umask
config:8-15

$ dmesg | grep "CFI failure"
[ 1258.174653] CFI failure (target: __uncore_umask_show...):

Update the type in the DEFINE_UNCORE_FORMAT_ATTR macro to match
'struct device_attribute' so that there is no more CFI violation.

Fixes: 06f2c24584f3 ("perf/amd/uncore: Prepare to scale for more attributes that vary per family")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210415001112.3024673-2-nathan@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agox86/events/amd/iommu: Fix sysfs type mismatch
Nathan Chancellor [Thu, 15 Apr 2021 00:11:11 +0000 (17:11 -0700)]
x86/events/amd/iommu: Fix sysfs type mismatch

[ Upstream commit de5bc7b425d4c27ae5faa00ea7eb6b9780b9a355 ]

dev_attr_show() calls _iommu_event_show() via an indirect call but
_iommu_event_show()'s type does not currently match the type of the
show() member in 'struct device_attribute', resulting in a Control Flow
Integrity violation.

$ cat /sys/devices/amd_iommu_1/events/mem_dte_hit
csource=0x0a

$ dmesg | grep "CFI failure"
[ 3526.735140] CFI failure (target: _iommu_event_show...):

Change _iommu_event_show() and 'struct amd_iommu_event_desc' to
'struct device_attribute' so that there is no more CFI violation.

Fixes: 7be6296fdd75 ("perf/x86/amd: AMD IOMMU Performance Counter PERF uncore PMU implementation")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210415001112.3024673-1-nathan@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoHSI: core: fix resource leaks in hsi_add_client_from_dt()
Dan Carpenter [Fri, 9 Apr 2021 11:08:17 +0000 (14:08 +0300)]
HSI: core: fix resource leaks in hsi_add_client_from_dt()

[ Upstream commit 5c08b0f75575648032f309a6f58294453423ed93 ]

If some of the allocations fail between the dev_set_name() and the
device_register() then the name will not be freed.  Fix this by
moving dev_set_name() directly in front of the call to device_register().

Fixes: a2aa24734d9d ("HSI: Add common DT binding for HSI client devices")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomedia: cedrus: Fix H265 status definitions
Jernej Skrabec [Mon, 12 Apr 2021 15:43:49 +0000 (17:43 +0200)]
media: cedrus: Fix H265 status definitions

[ Upstream commit 147d211cc9b4d753148d1640a1758b25edfbf437 ]

Some of the H265 status flags are wrong. Redefine them to corespond to
Allwinner CedarC open source userspace library. Only one of these flags
is actually used and new value also matches value used in libvdpau-sunxi
library, which is proven to be working.

Note that wrong (old) value in right circumstances (in combination with
another H265 decoding bug) causes driver lock up. With this fix decoding
is still broken (green output) but at least driver doesn't lock up.

Fixes: 86caab29da78 ("media: cedrus: Add HEVC/H.265 decoding support")
Signed-off-by: Jernej Skrabec <jernej.skrabec@siol.net>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonvme-pci: don't simple map sgl when sgls are disabled
Niklas Cassel [Fri, 9 Apr 2021 18:12:55 +0000 (20:12 +0200)]
nvme-pci: don't simple map sgl when sgls are disabled

[ Upstream commit e51183be1fa96dc6d3cd11b3c25a0f595807315e ]

According to the module parameter description for sgl_threshold,
a value of 0 means that SGLs are disabled.

If SGLs are disabled, we should respect that, even for the case
where the request is made up of a single physical segment.

Fixes: 297910571f08 ("nvme-pci: optimize mapping single segment requests using SGLs")
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agonvmet-tcp: fix a segmentation fault during io parsing error
Elad Grupi [Wed, 31 Mar 2021 09:13:14 +0000 (17:13 +0800)]
nvmet-tcp: fix a segmentation fault during io parsing error

[ Upstream commit bdaf13279192c60b2b1fc99badef53b494fec055 ]

In case there is an io that contains inline data and it goes to
parsing error flow, command response will free command and iov
before clearing the data on the socket buffer.
This will delay the command response until receive flow is completed.

Fixes: 872d26a391da ("nvmet-tcp: add NVMe over TCP target driver")
Signed-off-by: Elad Grupi <elad.grupi@dell.com>
Signed-off-by: Hou Pu <houpu.main@gmail.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomfd: stm32-timers: Avoid clearing auto reload register
Fabrice Gasnier [Wed, 3 Mar 2021 17:51:35 +0000 (18:51 +0100)]
mfd: stm32-timers: Avoid clearing auto reload register

[ Upstream commit 4917e498c6894ba077867aff78f82cffd5ffbb5c ]

The ARR register is cleared unconditionally upon probing, after the maximum
value has been read. This initial condition is rather not intuitive, when
considering the counter child driver. It rather expects the maximum value
by default:
- The counter interface shows a zero value by default for 'ceiling'
  attribute.
- Enabling the counter without any prior configuration makes it doesn't
  count.

The reset value of ARR register is the maximum. So Choice here
is to backup it, and restore it then, instead of clearing its value.
It also fixes the initial condition seen by the counter driver.

Fixes: d0f949e220fd ("mfd: Add STM32 Timers driver")
Signed-off-by: Fabrice Gasnier <fabrice.gasnier@foss.st.com>
Acked-by: William Breathitt Gray <vilhelm.gray@gmail.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agomailbox: sprd: Introduce refcnt when clients requests/free channels
Orson Zhai [Fri, 19 Mar 2021 06:15:35 +0000 (14:15 +0800)]
mailbox: sprd: Introduce refcnt when clients requests/free channels

[ Upstream commit 9468ab84032f96496e998cfa173cd1d0ac316bcd ]

Unisoc mailbox has no way to be enabled/disabled for any single channel.
They can only be set to startup or shutdown as a whole device at same time.

Add a variable to count references to avoid mailbox FIFO being reset
unexpectedly when clients are requesting or freeing channels.

Also add a lock to dismiss possible conflicts from register r/w in
different startup or shutdown threads. And fix the crash problem when early
interrupts come from channel which has not been requested by client yet.

Fixes: ca27fc26cd22 ("mailbox: sprd: Add Spreadtrum mailbox driver")
Signed-off-by: Orson Zhai <orson.zhai@unisoc.com>
Reviewed-by: Baolin Wang <baolin.wang7@gmail.com>
Signed-off-by: Jassi Brar <jaswinder.singh@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: ibmvfc: Fix invalid state machine BUG_ON()
Brian King [Tue, 13 Apr 2021 00:10:09 +0000 (18:10 -0600)]
scsi: ibmvfc: Fix invalid state machine BUG_ON()

[ Upstream commit 15cfef8623a449d40d16541687afd58e78033be3 ]

This fixes an issue hitting the BUG_ON() in ibmvfc_do_work(). When going
through a host action of IBMVFC_HOST_ACTION_RESET, we change the action to
IBMVFC_HOST_ACTION_TGT_DEL, then drop the host lock, and reset the CRQ,
which changes the host state to IBMVFC_NO_CRQ. If, prior to setting the
host state to IBMVFC_NO_CRQ, ibmvfc_init_host() is called, it can then end
up changing the host action to IBMVFC_HOST_ACTION_INIT.  If we then change
the host state to IBMVFC_NO_CRQ, we will then hit the BUG_ON().

Make a couple of changes to avoid this. Leave the host action to be
IBMVFC_HOST_ACTION_RESET or IBMVFC_HOST_ACTION_REENABLE until after we drop
the host lock and reset or reenable the CRQ. Also harden the host state
machine to ensure we cannot leave the reset / reenable state until we've
finished processing the reset or reenable.

Link: https://lore.kernel.org/r/20210413001009.902400-1-tyreld@linux.ibm.com
Fixes: 73ee5d867287 ("[SCSI] ibmvfc: Fix soft lockup on resume")
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
[tyreld: added fixes tag]
Signed-off-by: Tyrel Datwyler <tyreld@linux.ibm.com>
[mkp: fix comment checkpatch warnings]
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: sni_53c710: Add IRQ check
Sergey Shtylyov [Tue, 30 Mar 2021 17:45:12 +0000 (20:45 +0300)]
scsi: sni_53c710: Add IRQ check

[ Upstream commit 1160d61bc51e87e509cfaf9da50a0060f67b6de4 ]

The driver neglects to check the result of platform_get_irq()'s call and
blithely passes the negative error codes to request_irq() (which takes
*unsigned* IRQ #s), causing it to fail with -EINVAL (overridden by -ENODEV
further below).  Stop calling request_irq() with the invalid IRQ #s.

Link: https://lore.kernel.org/r/8f4b8fa5-8251-b977-70a1-9099bcb4bb17@omprussia.ru
Fixes: c27d85f3f3c5 ("[SCSI] SNI RM 53c710 driver")
Signed-off-by: Sergey Shtylyov <s.shtylyov@omprussia.ru>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: sun3x_esp: Add IRQ check
Sergey Shtylyov [Tue, 30 Mar 2021 17:44:08 +0000 (20:44 +0300)]
scsi: sun3x_esp: Add IRQ check

[ Upstream commit 14b321380eb333c82853d7d612d0995f05f88fdc ]

The driver neglects to check the result of platform_get_irq()'s call and
blithely passes the negative error codes to request_irq() (which takes
*unsigned* IRQ #), causing it to fail with -EINVAL, overriding the real
error code.  Stop calling request_irq() with the invalid IRQ #s.

Link: https://lore.kernel.org/r/363eb4c8-a3bf-4dc9-2a9e-90f349030a15@omprussia.ru
Fixes: 0bb67f181834 ("[SCSI] sun3x_esp: convert to esp_scsi")
Signed-off-by: Sergey Shtylyov <s.shtylyov@omprussia.ru>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: jazz_esp: Add IRQ check
Sergey Shtylyov [Tue, 30 Mar 2021 17:43:23 +0000 (20:43 +0300)]
scsi: jazz_esp: Add IRQ check

[ Upstream commit 38fca15c29db6ed06e894ac194502633e2a7d1fb ]

The driver neglects to check the result of platform_get_irq()'s call and
blithely passes the negative error codes to request_irq() (which takes
*unsigned* IRQ #), causing it to fail with -EINVAL, overriding the real
error code.  Stop calling request_irq() with the invalid IRQ #s.

Link: https://lore.kernel.org/r/594aa9ae-2215-49f6-f73c-33bd38989912@omprussia.ru
Fixes: 352e921f0dd4 ("[SCSI] jazz_esp: converted to use esp_core")
Signed-off-by: Sergey Shtylyov <s.shtylyov@omprussia.ru>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: hisi_sas: Fix IRQ checks
Sergey Shtylyov [Sat, 3 Apr 2021 20:43:55 +0000 (23:43 +0300)]
scsi: hisi_sas: Fix IRQ checks

[ Upstream commit 6c11dc060427e07ca144eacaccd696106b361b06 ]

Commit df2d8213d9e3 ("hisi_sas: use platform_get_irq()") failed to take
into account that irq_of_parse_and_map() and platform_get_irq() have a
different way of indicating an error: the former returns 0 and the latter
returns a negative error code. Fix up the IRQ checks!

Link: https://lore.kernel.org/r/810f26d3-908b-1d6b-dc5c-40019726baca@omprussia.ru
Fixes: df2d8213d9e3 ("hisi_sas: use platform_get_irq()")
Acked-by: John Garry <john.garry@huawei.com>
Signed-off-by: Sergey Shtylyov <s.shtylyov@omprussia.ru>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: ufs: ufshcd-pltfrm: Fix deferred probing
Sergey Shtylyov [Mon, 29 Mar 2021 20:50:58 +0000 (23:50 +0300)]
scsi: ufs: ufshcd-pltfrm: Fix deferred probing

[ Upstream commit 339c9b63cc7ce779ce45c675bf709cb58b807fc3 ]

The driver overrides the error codes returned by platform_get_irq() to
-ENODEV, so if it returns -EPROBE_DEFER, the driver would fail the probe
permanently instead of the deferred probing.  Propagate the error code
upstream as it should have been done from the start...

Link: https://lore.kernel.org/r/420364ca-614a-45e3-4e35-0e0653c7bc53@omprussia.ru
Fixes: 2953f850c3b8 ("[SCSI] ufs: use devres functions for ufshcd")
Signed-off-by: Sergey Shtylyov <s.shtylyov@omprussia.ru>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: pm80xx: Fix potential infinite loop
Colin Ian King [Wed, 7 Apr 2021 13:58:40 +0000 (14:58 +0100)]
scsi: pm80xx: Fix potential infinite loop

[ Upstream commit 40fa7394a1ad5706e795823276f2e394cca145d0 ]

The for-loop iterates with a u8 loop counter i and compares this with the
loop upper limit of pm8001_ha->max_q_num which is a u32 type.  There is a
potential infinite loop if pm8001_ha->max_q_num is larger than the u8 loop
counter. Fix this by making the loop counter the same type as
pm8001_ha->max_q_num.

[mkp: this is purely theoretical, max_q_num is currently limited to 64]

Link: https://lore.kernel.org/r/20210407135840.494747-1-colin.king@canonical.com
Fixes: 65df7d1986a1 ("scsi: pm80xx: Fix chip initialization failure")
Addresses-Coverity: ("Infinite loop")
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoscsi: pm80xx: Increase timeout for pm80xx mpi_uninit_check()
Igor Pylypiv [Tue, 6 Apr 2021 18:05:33 +0000 (11:05 -0700)]
scsi: pm80xx: Increase timeout for pm80xx mpi_uninit_check()

[ Upstream commit 3f744a14f331f56703a9d74e86520db045f11831 ]

The mpi_uninit_check() takes longer for inbound doorbell register to be
cleared. Increase the timeout substantially so that the driver does not
fail to load.

Previously, the inbound doorbell wait time was mistakenly increased in the
mpi_init_check() instead of mpi_uninit_check(). It is okay to leave the
mpi_init_check() wait time as-is as these are timeout values and if there
is a failure, waiting longer is not an issue.

Link: https://lore.kernel.org/r/20210406180534.1924345-2-ipylypiv@google.com
Fixes: e90e236250e9 ("scsi: pm80xx: Increase timeout for pm80xx mpi_uninit_check")
Reviewed-by: Vishakha Channapattan <vishakhavc@google.com>
Acked-by: Jack Wang <jinpu.wang@ionos.com>
Signed-off-by: Igor Pylypiv <ipylypiv@google.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoclk: uniphier: Fix potential infinite loop
Colin Ian King [Fri, 9 Apr 2021 09:01:03 +0000 (10:01 +0100)]
clk: uniphier: Fix potential infinite loop

[ Upstream commit f6b1340dc751a6caa2a0567b667d0f4f4172cd58 ]

The for-loop iterates with a u8 loop counter i and compares this
with the loop upper limit of num_parents that is an int type.
There is a potential infinite loop if num_parents is larger than
the u8 loop counter. Fix this by making the loop counter the same
type as num_parents.  Also make num_parents an unsigned int to
match the return type of the call to clk_hw_get_num_parents.

Addresses-Coverity: ("Infinite loop")
Fixes: 734d82f4a678 ("clk: uniphier: add core support code for UniPhier clock driver")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Reviewed-by: Masahiro Yamada <masahiroy@kernel.org>
Link: https://lore.kernel.org/r/20210409090104.629722-1-colin.king@canonical.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agodrm/radeon: Fix a missing check bug in radeon_dp_mst_detect()
Yingjie Wang [Wed, 7 Apr 2021 03:10:04 +0000 (20:10 -0700)]
drm/radeon: Fix a missing check bug in radeon_dp_mst_detect()

[ Upstream commit 25315ebfaefcffd126a266116b37bb8a3d1c4620 ]

In radeon_dp_mst_detect(), We should check whether or not @connector
has been unregistered from userspace. If the connector is unregistered,
we should return disconnected status.

Fixes: 9843ead08f18 ("drm/radeon: add DisplayPort MST support (v2)")
Signed-off-by: Yingjie Wang <wangyingjie55@126.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agodrm/amd/display: use GFP_ATOMIC in dcn20_resource_construct
Nirmoy Das [Wed, 17 Mar 2021 10:38:11 +0000 (11:38 +0100)]
drm/amd/display: use GFP_ATOMIC in dcn20_resource_construct

[ Upstream commit 3bb1105071fb974e3e3ca2f92ddfd69c81285ab6 ]

Replace GFP_KERNEL with GFP_ATOMIC as dcn20_resource_construct()
can't sleep.

Partially fixes: https://bugzilla.kernel.org/show_bug.cgi?id=212311
as dcn20_resource_construct() also calls into SMU functions which does
mutex_lock().

Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoclk: qcom: apss-ipq-pll: Add missing MODULE_DEVICE_TABLE
Chen Hui [Fri, 9 Apr 2021 08:23:52 +0000 (16:23 +0800)]
clk: qcom: apss-ipq-pll: Add missing MODULE_DEVICE_TABLE

[ Upstream commit d0a859edda46b45baeab9687d173102300d76e2b ]

CONFIG_IPQ_APSS_PLL is tristate option and therefore this driver can
be compiled as a module. This patch adds missing MODULE_DEVICE_TABLE
definition which generates correct modalias for automatic loading of
this driver when it is built as an external module.

Fixes: ecd2bacfbbc4 ("clk: qcom: Add ipq apss pll driver")
Signed-off-by: Chen Hui <clare.chenhui@huawei.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Link: https://lore.kernel.org/r/20210409082352.233810-4-clare.chenhui@huawei.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoclk: qcom: a53-pll: Add missing MODULE_DEVICE_TABLE
Chen Hui [Fri, 9 Apr 2021 08:23:51 +0000 (16:23 +0800)]
clk: qcom: a53-pll: Add missing MODULE_DEVICE_TABLE

[ Upstream commit 790b516ada10a4dcc0f0a56dc0ced475d86d5820 ]

CONFIG_QCOM_A53PLL is tristate option and therefore this driver can be
compiled as a module. This patch adds missing MODULE_DEVICE_TABLE
definition which generates correct modalias for automatic loading of
this driver when it is built as an external module.

Fixes: 0c6ab1b8f894 ("clk: qcom: Add A53 PLL support")
Signed-off-by: Chen Hui <clare.chenhui@huawei.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Link: https://lore.kernel.org/r/20210409082352.233810-3-clare.chenhui@huawei.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agodrm: xlnx: zynqmp: fix a memset in zynqmp_dp_train()
Dan Carpenter [Tue, 30 Mar 2021 09:31:52 +0000 (12:31 +0300)]
drm: xlnx: zynqmp: fix a memset in zynqmp_dp_train()

[ Upstream commit 5842ab76bbfadb37eaea91e53c1efe34ae504e4a ]

The dp->train_set[] for this driver is only two characters, not four so
this memsets too much.  Fortunately, this ends up corrupting a struct
hole and not anything important.

Fixes: d76271d22694 ("drm: xlnx: DRM/KMS driver for Xilinx ZynqMP DisplayPort Subsystem")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/YGLwCBMotnrKZu6P@mwanda
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoclk: zynqmp: pll: add set_pll_mode to check condition in zynqmp_pll_enable
Quanyang Wang [Tue, 6 Apr 2021 15:31:31 +0000 (23:31 +0800)]
clk: zynqmp: pll: add set_pll_mode to check condition in zynqmp_pll_enable

[ Upstream commit 394cdb69a3c30b33524cf1204afe5cceaba69cdc ]

If there is a IOCTL_SET_PLL_FRAC_MODE request sent to ATF ever,
we shouldn't skip invoking PM_CLOCK_ENABLE fn even though this
pll has been enabled. In ATF implementation, it will only assign
the mode to the variable (struct pm_pll *)pll->mode when handling
IOCTL_SET_PLL_FRAC_MODE call. Invoking PM_CLOCK_ENABLE can force
ATF send request to PWU to set the pll mode to PLL's register.

There is a scenario that happens in enabling VPLL_INT(clk_id:96):
1) VPLL_INT has been enabled during booting.
2) A driver calls clk_set_rate and according to the rate, the VPLL_INT
   should be set to FRAC mode. Then zynqmp_pll_set_mode is called
   to pass IOCTL_SET_PLL_FRAC_MODE to ATF. Note that at this point
   ATF just stores the mode to a variable.
3) This driver calls clk_prepare_enable and zynqmp_pll_enable is
   called to try to enable VPLL_INT pll. Because of 1), the function
   zynqmp_pll_enable just returns without doing anything after checking
   that this pll has been enabled.

In the scenario above, the pll mode of VPLL_INT will never be set
successfully. So adding set_pll_mode to check condition to fix it.

Fixes: 3fde0e16d016 ("drivers: clk: Add ZynqMP clock driver")
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Tested-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Link: https://lore.kernel.org/r/20210406153131.601701-1-quanyang.wang@windriver.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
3 years agoclk: zynqmp: move zynqmp_pll_set_mode out of round_rate callback
Quanyang Wang [Tue, 6 Apr 2021 15:40:15 +0000 (23:40 +0800)]
clk: zynqmp: move zynqmp_pll_set_mode out of round_rate callback

[ Upstream commit d7fd3f9f53df8bb2212dff70f66f12cae0e1a653 ]

The round_rate callback should only perform rate calculation and not
involve calling zynqmp_pll_set_mode to change the pll mode. So let's
move zynqmp_pll_set_mode out of round_rate and to set_rate callback.

Fixes: 3fde0e16d016 ("drivers: clk: Add ZynqMP clock driver")
Reported-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Quanyang Wang <quanyang.wang@windriver.com>
Link: https://lore.kernel.org/r/20210406154015.602779-1-quanyang.wang@windriver.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>