platform/kernel/linux-rpi.git
23 months agohwrng: core - treat default_quality as a maximum and default to 1024
Jason A. Donenfeld [Mon, 7 Nov 2022 12:24:55 +0000 (13:24 +0100)]
hwrng: core - treat default_quality as a maximum and default to 1024

Most hw_random devices return entropy which is assumed to be of full
quality, but driver authors don't bother setting the quality knob. Some
hw_random devices return less than full quality entropy, and then driver
authors set the quality knob. Therefore, the entropy crediting should be
opt-out rather than opt-in per-driver, to reflect the actual reality on
the ground.

For example, the two Raspberry Pi RNG drivers produce full entropy
randomness, and both EDK2 and U-Boot's drivers for these treat them as
such. The result is that EFI then uses these numbers and passes the to
Linux, and Linux credits them as boot, thereby initializing the RNG.
Yet, in Linux, the quality knob was never set to anything, and so on the
chance that Linux is booted without EFI, nothing is ever credited.
That's annoying.

The same pattern appears to repeat itself throughout various drivers. In
fact, very very few drivers have bothered setting quality=1024.

Looking at the git history of existing drivers and corresponding mailing
list discussion, this conclusion tracks. There's been a decent amount of
discussion about drivers that set quality < 1024 -- somebody read and
interepreted a datasheet, or made some back of the envelope calculation
somehow. But there's been very little, if any, discussion about most
drivers where the quality is just set to 1024 or unset (or set to 1000
when the authors misunderstood the API and assumed it was base-10 rather
than base-2); in both cases the intent was fairly clear of, "this is a
hardware random device; it's fine."

So let's invert this logic. A hw_random struct's quality knob now
controls the maximum quality a driver can produce, or 0 to specify 1024.
Then, the module-wide switch called "default_quality" is changed to
represent the maximum quality of any driver. By default it's 1024, and
the quality of any particular driver is then given by:

    min(default_quality, rng->quality ?: 1024);

This way, the user can still turn this off for weird reasons (and we can
replace whatever driver-specific disabling hacks existed in the past),
yet we get proper crediting for relevant RNGs.

Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: qat - remove ADF_STATUS_PF_RUNNING flag from probe
Shashank Gupta [Fri, 4 Nov 2022 17:21:07 +0000 (13:21 -0400)]
crypto: qat - remove ADF_STATUS_PF_RUNNING flag from probe

The ADF_STATUS_PF_RUNNING bit is set after the successful initialization
of the communication between VF to PF in adf_vf2pf_notify_init().
So, it is not required to be set after the execution of the function
adf_dev_init().

Signed-off-by: Shashank Gupta <shashank.gupta@intel.com>
Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Reviewed-by: Wojciech Ziemba <wojciech.ziemba@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - Remove surplus dev_err() when using platform_get_irq()
Yang Li [Fri, 4 Nov 2022 07:45:27 +0000 (15:45 +0800)]
crypto: rockchip - Remove surplus dev_err() when using platform_get_irq()

There is no need to call the dev_err() function directly to print a
custom message when handling an error from either the platform_get_irq()
or platform_get_irq_byname() functions as both are going to display an
appropriate error message in case of a failure.

./drivers/crypto/rockchip/rk3288_crypto.c:351:2-9: line 351 is
redundant because platform_get_irq() already prints an error

Link: https://bugzilla.openanolis.cn/show_bug.cgi?id=2677
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Acked-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: lib/aesgcm - Provide minimal library implementation
Ard Biesheuvel [Thu, 3 Nov 2022 19:22:59 +0000 (20:22 +0100)]
crypto: lib/aesgcm - Provide minimal library implementation

Implement a minimal library version of AES-GCM based on the existing
library implementations of AES and multiplication in GF(2^128). Using
these primitives, GCM can be implemented in a straight-forward manner.

GCM has a couple of sharp edges, i.e., the amount of input data
processed with the same initialization vector (IV) should be capped to
protect the counter from 32-bit rollover (or carry), and the size of the
authentication tag should be fixed for a given key. [0]

The former concern is addressed trivially, given that the function call
API uses 32-bit signed types for the input lengths. It is still up to
the caller to avoid IV reuse in general, but this is not something we
can police at the implementation level.

As for the latter concern, let's make the authentication tag size part
of the key schedule, and only permit it to be configured as part of the
key expansion routine.

Note that table based AES implementations are susceptible to known
plaintext timing attacks on the encryption key. The AES library already
attempts to mitigate this to some extent, but given that the counter
mode encryption used by GCM operates exclusively on known plaintext by
construction (the IV and therefore the initial counter value are known
to an attacker), let's take some extra care to mitigate this, by calling
the AES library with interrupts disabled.

[0] https://nvlpubs.nist.gov/nistpubs/legacy/sp/nistspecialpublication800-38d.pdf

Link: https://lore.kernel.org/all/c6fb9b25-a4b6-2e4a-2dd1-63adda055a49@amd.com/
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Tested-by: Nikunj A Dadhania <nikunj@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: lib/gf128mul - make gf128mul_lle time invariant
Ard Biesheuvel [Thu, 3 Nov 2022 19:22:58 +0000 (20:22 +0100)]
crypto: lib/gf128mul - make gf128mul_lle time invariant

The gf128mul library has different variants with different
memory/performance tradeoffs, where the faster ones use 4k or 64k lookup
tables precomputed at runtime, which are based on one of the
multiplication factors, which is commonly the key for keyed hash
algorithms such as GHASH.

The slowest variant is gf128_mul_lle() [and its bbe/ble counterparts],
which does not use precomputed lookup tables, but it still relies on a
single u16[256] lookup table which is input independent. The use of such
a table may cause the execution time of gf128_mul_lle() to correlate
with the value of the inputs, which is generally something that must be
avoided for cryptographic algorithms. On top of that, the function uses
a sequence of if () statements that conditionally invoke be128_xor()
based on which bits are set in the second argument of the function,
which is usually a pointer to the multiplication factor that represents
the key.

In order to remove the correlation between the execution time of
gf128_mul_lle() and the value of its inputs, let's address the
identified shortcomings:
- add a time invariant version of gf128mul_x8_lle() that replaces the
  table lookup with the expression that is used at compile time to
  populate the lookup table;
- make the invocations of be128_xor() unconditional, but pass a zero
  vector as the third argument if the associated bit in the key is
  cleared.

The resulting code is likely to be significantly slower. However, given
that this is the slowest version already, making it even slower in order
to make it more secure is assumed to be justified.

The bbe and ble counterparts could receive the same treatment, but the
former is never used anywhere in the kernel, and the latter is only
used in the driver for a asynchronous crypto h/w accelerator (Chelsio),
where timing variances are unlikely to matter.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: move gf128mul library into lib/crypto
Ard Biesheuvel [Thu, 3 Nov 2022 19:22:57 +0000 (20:22 +0100)]
crypto: move gf128mul library into lib/crypto

The gf128mul library does not depend on the crypto API at all, so it can
be moved into lib/crypto. This will allow us to use it in other library
code in a subsequent patch without having to depend on CONFIG_CRYPTO.

While at it, change the Kconfig symbol name to align with other crypto
library implementations. However, the source file name is retained, as
it is reflected in the module .ko filename, and changing this might
break things for users.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: doc - use correct function name
Ralph Siemsen [Thu, 27 Oct 2022 19:35:44 +0000 (15:35 -0400)]
crypto: doc - use correct function name

The hashing API does not have a function called .finish()

Signed-off-by: Ralph Siemsen <ralph.siemsen@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - add CE implementation for GCM mode
Tianjia Zhang [Thu, 27 Oct 2022 06:55:05 +0000 (14:55 +0800)]
crypto: arm64/sm4 - add CE implementation for GCM mode

This patch is a CE-optimized assembly implementation for GCM mode.

Benchmark on T-Head Yitian-710 2.75 GHz, the data comes from the 224 and 224
modes of tcrypt, and compared the performance before and after this patch (the
driver used before this patch is gcm_base(ctr-sm4-ce,ghash-generic)).
The abscissas are blocks of different lengths. The data is tabulated and the
unit is Mb/s:

Before (gcm_base(ctr-sm4-ce,ghash-generic)):

gcm(sm4)     |     16      64      256      512     1024     1420     4096     8192
-------------+---------------------------------------------------------------------
  GCM enc    |  25.24   64.65   104.66   116.69   123.81   125.12   129.67   130.62
  GCM dec    |  25.40   64.80   104.74   116.70   123.81   125.21   129.68   130.59
  GCM mb enc |  24.95   64.06   104.20   116.38   123.55   124.97   129.63   130.61
  GCM mb dec |  24.92   64.00   104.13   116.34   123.55   124.98   129.56   130.48

After:

gcm-sm4-ce   |     16      64      256      512     1024     1420     4096     8192
-------------+---------------------------------------------------------------------
  GCM enc    | 108.62  397.18   971.60  1283.92  1522.77  1513.39  1777.00  1806.96
  GCM dec    | 116.36  398.14  1004.27  1319.11  1624.21  1635.43  1932.54  1974.20
  GCM mb enc | 107.13  391.79   962.05  1274.94  1514.76  1508.57  1769.07  1801.58
  GCM mb dec | 113.40  389.36   988.51  1307.68  1619.10  1631.55  1931.70  1970.86

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - add CE implementation for CCM mode
Tianjia Zhang [Thu, 27 Oct 2022 06:55:04 +0000 (14:55 +0800)]
crypto: arm64/sm4 - add CE implementation for CCM mode

This patch is a CE-optimized assembly implementation for CCM mode.

Benchmark on T-Head Yitian-710 2.75 GHz, the data comes from the 223 and 225
modes of tcrypt, and compared the performance before and after this patch (the
driver used before this patch is ccm_base(ctr-sm4-ce,cbcmac-sm4-ce)).
The abscissas are blocks of different lengths. The data is tabulated and the
unit is Mb/s:

Before (rfc4309(ccm_base(ctr-sm4-ce,cbcmac-sm4-ce))):

ccm(sm4)     |     16      64     256     512    1024    1420    4096    8192
-------------+---------------------------------------------------------------
  CCM enc    |  35.07  125.40  336.47  468.17  581.97  619.18  712.56  736.01
  CCM dec    |  34.87  124.40  335.08  466.75  581.04  618.81  712.25  735.89
  CCM mb enc |  34.71  123.96  333.92  465.39  579.91  617.49  711.45  734.92
  CCM mb dec |  34.42  122.80  331.02  462.81  578.28  616.42  709.88  734.19

After (rfc4309(ccm-sm4-ce)):

ccm-sm4-ce   |     16      64     256     512    1024    1420    4096    8192
-------------+---------------------------------------------------------------
  CCM enc    |  77.12  249.82  569.94  725.17  839.27  867.71  952.87  969.89
  CCM dec    |  75.90  247.26  566.29  722.12  836.90  865.95  951.74  968.57
  CCM mb enc |  75.98  245.25  562.91  718.99  834.76  864.70  950.17  967.90
  CCM mb dec |  75.06  243.78  560.58  717.13  833.68  862.70  949.35  967.11

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - add CE implementation for cmac/xcbc/cbcmac
Tianjia Zhang [Thu, 27 Oct 2022 06:55:03 +0000 (14:55 +0800)]
crypto: arm64/sm4 - add CE implementation for cmac/xcbc/cbcmac

This patch is a CE-optimized assembly implementation for cmac/xcbc/cbcmac.

Benchmark on T-Head Yitian-710 2.75 GHz, the data comes from the 300 mode of
tcrypt, and compared the performance before and after this patch (the driver
used before this patch is XXXmac(sm4-ce)). The abscissas are blocks of
different lengths. The data is tabulated and the unit is Mb/s:

Before:

update-size    |      16      64     256    1024    2048    4096    8192
---------------+--------------------------------------------------------
cmac(sm4-ce)   |  293.33  403.69  503.76  527.78  531.10  535.46  535.81
xcbc(sm4-ce)   |  292.83  402.50  504.02  529.08  529.87  536.55  538.24
cbcmac(sm4-ce) |  318.42  415.79  497.12  515.05  523.15  521.19  523.01

After:

update-size    |      16      64     256    1024    2048    4096    8192
---------------+--------------------------------------------------------
cmac-sm4-ce    |  371.99  675.28  903.56  971.65  980.57  990.40  991.04
xcbc-sm4-ce    |  372.11  674.55  903.47  971.61  980.96  990.42  991.10
cbcmac-sm4-ce  |  371.63  675.33  903.23  972.07  981.42  990.93  991.45

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - add CE implementation for XTS mode
Tianjia Zhang [Thu, 27 Oct 2022 06:55:02 +0000 (14:55 +0800)]
crypto: arm64/sm4 - add CE implementation for XTS mode

This patch is a CE-optimized assembly implementation for XTS mode.

Benchmark on T-Head Yitian-710 2.75 GHz, the data comes from the 218 mode of
tcrypt, and compared the performance before and after this patch (the driver
used before this patch is xts(ecb-sm4-ce)). The abscissas are blocks of
different lengths. The data is tabulated and the unit is Mb/s:

Before:

xts(ecb-sm4-ce) |      16       64      128      256     1024     1420     4096
----------------+--------------------------------------------------------------
        XTS enc |  117.17   430.56   732.92  1134.98  2007.03  2136.23  2347.20
        XTS dec |  116.89   429.02   733.40  1132.96  2006.13  2130.50  2347.92

After:

xts-sm4-ce      |      16       64      128      256     1024     1420     4096
----------------+--------------------------------------------------------------
        XTS enc |  224.68   798.91  1248.08  1714.60  2413.73  2467.84  2612.62
        XTS dec |  229.85   791.34  1237.79  1720.00  2413.30  2473.84  2611.95

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - add CE implementation for CTS-CBC mode
Tianjia Zhang [Thu, 27 Oct 2022 06:55:01 +0000 (14:55 +0800)]
crypto: arm64/sm4 - add CE implementation for CTS-CBC mode

This patch is a CE-optimized assembly implementation for CTS-CBC mode.

Benchmark on T-Head Yitian-710 2.75 GHz, the data comes from the 218 mode of
tcrypt, and compared the performance before and after this patch (the driver
used before this patch is cts(cbc-sm4-ce)). The abscissas are blocks of
different lengths. The data is tabulated and the unit is Mb/s:

Before:

cts(cbc-sm4-ce) |      16       64      128      256     1024     1420     4096
----------------+--------------------------------------------------------------
    CTS-CBC enc |  286.09   297.17   457.97   627.75   868.58   900.80   957.69
    CTS-CBC dec |  286.67   285.63   538.35   947.08  2241.03  2577.32  3391.14

After:

cts-cbc-sm4-ce  |      16       64      128      256     1024     1420     4096
----------------+--------------------------------------------------------------
    CTS-CBC enc |  288.19   428.80   593.57   741.04   911.73   931.80   950.00
    CTS-CBC dec |  292.22   468.99   838.23  1380.76  2741.17  3036.42  3409.62

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - export reusable CE acceleration functions
Tianjia Zhang [Thu, 27 Oct 2022 06:55:00 +0000 (14:55 +0800)]
crypto: arm64/sm4 - export reusable CE acceleration functions

In the accelerated implementation of the SM4 algorithm using the Crypto
Extension instructions, there are some functions that can be reused in
the upcoming accelerated implementation of the GCM/CCM mode, and the
CBC/CFB encryption is reused in the optimized implementation of SVESM4.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - simplify sm4_ce_expand_key() of CE implementation
Tianjia Zhang [Thu, 27 Oct 2022 06:54:59 +0000 (14:54 +0800)]
crypto: arm64/sm4 - simplify sm4_ce_expand_key() of CE implementation

Use a 128-bit swap mask and tbl instruction to simplify the implementation
for generating SM4 rkey_dec.

Also fixed the issue of not being wrapped by kernel_neon_begin/end() when
using the sm4_ce_expand_key() function.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - refactor and simplify CE implementation
Tianjia Zhang [Thu, 27 Oct 2022 06:54:58 +0000 (14:54 +0800)]
crypto: arm64/sm4 - refactor and simplify CE implementation

This patch does not add new features, but only refactors and simplifies the
implementation of the Crypto Extension acceleration of the SM4 algorithm:

Extract the macro optimized by SM4 Crypto Extension for reuse in the
subsequent optimization of CCM/GCM modes.

Encryption in CBC and CFB modes processes four blocks at a time instead of
one, allowing the ld1 instruction to load 64 bytes of data at a time, which
will reduces unnecessary memory accesses.

CBC/CFB/CTR makes full use of free registers to reduce redundant memory
accesses, and rearranges some instructions to improve out-of-order execution
capabilities.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: tcrypt - add SM4 cts-cbc/xts/xcbc test
Tianjia Zhang [Thu, 27 Oct 2022 06:54:57 +0000 (14:54 +0800)]
crypto: tcrypt - add SM4 cts-cbc/xts/xcbc test

Added CTS-CBC/XTS/XCBC tests for SM4 algorithms, as well as
corresponding speed tests, this is to test performance-optimized
implementations of these modes.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: testmgr - add SM4 cts-cbc/xts/xcbc test vectors
Tianjia Zhang [Thu, 27 Oct 2022 06:54:56 +0000 (14:54 +0800)]
crypto: testmgr - add SM4 cts-cbc/xts/xcbc test vectors

This patch newly adds the test vectors of CTS-CBC/XTS/XCBC modes of
the SM4 algorithm, and also added some test vectors for SM4 GCM/CCM.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm4 - refactor and simplify NEON implementation
Tianjia Zhang [Thu, 27 Oct 2022 06:54:55 +0000 (14:54 +0800)]
crypto: arm64/sm4 - refactor and simplify NEON implementation

This patch does not add new features. The main work is to refactor and
simplify the implementation of SM4 NEON, which is reflected in the
following aspects:

The accelerated implementation supports the arbitrary number of blocks,
not just multiples of 8, which simplifies the implementation and brings
some optimization acceleration for data that is not aligned by 8 blocks.

When loading the input data, use the ld4 instruction to replace the
original ld1 instruction as much as possible, which will save the cost
of matrix transposition of the input data.

Use 8-block parallelism whenever possible to speed up matrix transpose
and rotation operations, instead of up to 4-block parallelism.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm3 - add NEON assembly implementation
Tianjia Zhang [Thu, 27 Oct 2022 06:54:54 +0000 (14:54 +0800)]
crypto: arm64/sm3 - add NEON assembly implementation

This patch adds the NEON acceleration implementation of the SM3 hash
algorithm. The main algorithm is based on SM3 NEON accelerated work of
the libgcrypt project.

Benchmark on T-Head Yitian-710 2.75 GHz, the data comes from the 326 mode
of tcrypt, and compares the performance data of sm3-generic and sm3-ce.
The abscissas are blocks of different lengths. The data is tabulated and
the unit is Mb/s:

update-size    |      16      64     256    1024    2048    4096    8192
---------------+--------------------------------------------------------
sm3-generic    |  185.24  221.28  301.26  307.43  300.83  308.82  308.91
sm3-neon       |  171.81  220.20  322.94  339.28  334.09  343.61  343.87
sm3-ce         |  227.48  333.48  502.62  527.87  520.45  534.91  535.40

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: arm64/sm3 - raise the priority of the CE implementation
Tianjia Zhang [Thu, 27 Oct 2022 06:54:53 +0000 (14:54 +0800)]
crypto: arm64/sm3 - raise the priority of the CE implementation

Raise the priority of the sm3-ce algorithm from 200 to 400, this is
to make room for the implementation of sm3-neon.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: tcrypt - Drop leading newlines from prints
Anirudh Venkataramanan [Wed, 26 Oct 2022 19:16:16 +0000 (12:16 -0700)]
crypto: tcrypt - Drop leading newlines from prints

The top level print banners have a leading newline. It's not entirely
clear why this exists, but it makes it harder to parse tcrypt test output
using a script. Drop said newlines.

tcrypt output before this patch:

[...]
      testing speed of rfc4106(gcm(aes)) (rfc4106-gcm-aesni) encryption
[...] test 0 (160 bit key, 16 byte blocks): 1 operation in 2320 cycles (16 bytes)

tcrypt output with this patch:

[...] testing speed of rfc4106(gcm(aes)) (rfc4106-gcm-aesni) encryption
[...] test 0 (160 bit key, 16 byte blocks): 1 operation in 2320 cycles (16 bytes)

Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: tcrypt - Drop module name from print string
Anirudh Venkataramanan [Wed, 26 Oct 2022 19:16:15 +0000 (12:16 -0700)]
crypto: tcrypt - Drop module name from print string

The pr_fmt() define includes KBUILD_MODNAME, and so there's no need
for pr_err() to also print it. Drop module name from the print string.

Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: tcrypt - Use pr_info/pr_err
Anirudh Venkataramanan [Wed, 26 Oct 2022 19:16:14 +0000 (12:16 -0700)]
crypto: tcrypt - Use pr_info/pr_err

Currently, there's mixed use of printk() and pr_info()/pr_err(). The latter
prints the module name (because pr_fmt() is defined so) but the former does
not. As a result there's inconsistency in the printed output. For example:

modprobe mode=211:

[...] test 0 (160 bit key, 16 byte blocks): 1 operation in 2320 cycles (16 bytes)
[...] test 1 (160 bit key, 64 byte blocks): 1 operation in 2336 cycles (64 bytes)

modprobe mode=215:

[...] tcrypt: test 0 (160 bit key, 16 byte blocks): 1 operation in 2173 cycles (16 bytes)
[...] tcrypt: test 1 (160 bit key, 64 byte blocks): 1 operation in 2241 cycles (64 bytes)

Replace all instances of printk() with pr_info()/pr_err() so that the
module name is printed consistently.

Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: tcrypt - Use pr_cont to print test results
Anirudh Venkataramanan [Wed, 26 Oct 2022 19:16:13 +0000 (12:16 -0700)]
crypto: tcrypt - Use pr_cont to print test results

For some test cases, a line break gets inserted between the test banner
and the results. For example, with mode=211 this is the output:

[...]
      testing speed of rfc4106(gcm(aes)) (rfc4106-gcm-aesni) encryption
[...] test 0 (160 bit key, 16 byte blocks):
[...] 1 operation in 2373 cycles (16 bytes)

--snip--

[...]
      testing speed of gcm(aes) (generic-gcm-aesni) encryption
[...] test 0 (128 bit key, 16 byte blocks):
[...] 1 operation in 2338 cycles (16 bytes)

Similar behavior is seen in the following cases as well:

  modprobe tcrypt mode=212
  modprobe tcrypt mode=213
  modprobe tcrypt mode=221
  modprobe tcrypt mode=300 sec=1
  modprobe tcrypt mode=400 sec=1

This doesn't happen with mode=215:

[...] tcrypt:
              testing speed of multibuffer rfc4106(gcm(aes)) (rfc4106-gcm-aesni) encryption
[...] tcrypt: test 0 (160 bit key, 16 byte blocks): 1 operation in 2215 cycles (16 bytes)

--snip--

[...] tcrypt:
              testing speed of multibuffer gcm(aes) (generic-gcm-aesni) encryption
[...] tcrypt: test 0 (128 bit key, 16 byte blocks): 1 operation in 2191 cycles (16 bytes)

This print inconsistency is because printk() is used instead of pr_cont()
in a few places. Change these to be pr_cont().

checkpatch warns that pr_cont() shouldn't be used. This can be ignored in
this context as tcrypt already uses pr_cont().

Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: octeontx - fix repeated words in comments
wangjianli [Sat, 22 Oct 2022 05:38:02 +0000 (13:38 +0800)]
crypto: octeontx - fix repeated words in comments

Delete the redundant word 'the'.

Signed-off-by: wangjianli <wangjianli@cdjrlc.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/qm - delete redundancy check
Kai Ye [Sat, 22 Oct 2022 01:17:46 +0000 (01:17 +0000)]
crypto: hisilicon/qm - delete redundancy check

Because the permission on the VF debugfs file is "0444". So
the VF function checking is redundant in qos writing api.

Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/qm - add pci bdf number check
Kai Ye [Sat, 22 Oct 2022 01:17:45 +0000 (01:17 +0000)]
crypto: hisilicon/qm - add pci bdf number check

The pci bdf number check is added for qos written by using the pci api.
Directly get the devfn by pci_dev, so delete some redundant code.
And use the kstrtoul instead of sscanf to simplify code.

Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/qm - increase the memory of local variables
Kai Ye [Sat, 22 Oct 2022 01:17:44 +0000 (01:17 +0000)]
crypto: hisilicon/qm - increase the memory of local variables

Increase the buffer to prevent stack overflow by fuzz test. The maximum
length of the qos configuration buffer is 256 bytes. Currently, the value
of the 'val buffer' is only 32 bytes. The sscanf does not check the dest
memory length. So the 'val buffer' may stack overflow.

Signed-off-by: Kai Ye <yekai13@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: af_alg - Support symmetric encryption via keyring keys
Frederick Lawler [Mon, 17 Oct 2022 19:25:00 +0000 (14:25 -0500)]
crypto: af_alg - Support symmetric encryption via keyring keys

We want to leverage keyring to store sensitive keys, and then use those
keys for symmetric encryption via the crypto API. Among the key types we
wish to support are: user, logon, encrypted, and trusted.

User key types are already able to have their data copied to user space,
but logon does not support this. Further, trusted and encrypted keys will
return their encrypted data back to user space on read, which does not
make them ideal for symmetric encryption.

To support symmetric encryption for these key types, add a new
ALG_SET_KEY_BY_KEY_SERIAL setsockopt() option to the crypto API. This
allows users to pass a key_serial_t to the crypto API to perform
symmetric encryption. The behavior is the same as ALG_SET_KEY, but
the crypto key data is copied in kernel space from a keyring key,
which allows for the support of logon, encrypted, and trusted key types.

Keyring keys must have the KEY_(POS|USR|GRP|OTH)_SEARCH permission set
to leverage this feature. This follows the asymmetric_key type where key
lookup calls eventually lead to keyring_search_rcu() without the
KEYRING_SEARCH_NO_CHECK_PERM flag set.

Signed-off-by: Frederick Lawler <fred@cloudflare.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - Add support for RK3399
Corentin Labbe [Tue, 27 Sep 2022 07:55:11 +0000 (07:55 +0000)]
crypto: rockchip - Add support for RK3399

The RK3399 has 2 rk3288 compatible crypto device named crypto0 and
crypto1. The only difference is lack of RSA in crypto1.

We need to add driver support for 2 parallel instance as only one need
to register crypto algorithms.
Then the driver will round robin each request on each device.

For avoiding complexity (device bringup after a TFM is created), PM is
modified to be handled per request.
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - permit to have more than one reset
Corentin Labbe [Tue, 27 Sep 2022 07:55:10 +0000 (07:55 +0000)]
crypto: rockchip - permit to have more than one reset

The RK3399 has 3 resets, so the driver to handle multiple resets.
This is done by using devm_reset_control_array_get_exclusive().

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - rk_ahash_reg_init use crypto_info from parameter
Corentin Labbe [Tue, 27 Sep 2022 07:55:09 +0000 (07:55 +0000)]
crypto: rockchip - rk_ahash_reg_init use crypto_info from parameter

rk_ahash_reg_init() use crypto_info from TFM context, since we will
remove it, let's take if from parameters.

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - Check for clocks numbers and their frequencies
Corentin Labbe [Tue, 27 Sep 2022 07:55:08 +0000 (07:55 +0000)]
crypto: rockchip - Check for clocks numbers and their frequencies

Add the number of clocks needed for each compatible.
Rockchip's datasheet give maximum frequencies for some clocks, so add
checks for verifying they are within limits. Let's start with rk3288 for
clock frequency check, other will came later.

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - store crypto_info in request context
Corentin Labbe [Tue, 27 Sep 2022 07:55:07 +0000 (07:55 +0000)]
crypto: rockchip - store crypto_info in request context

The crypto_info to use must be stored in the request context.
This will help when 2 crypto_info will be available on rk3399.

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agodt-bindings: crypto: rockchip: add new compatible
Corentin Labbe [Tue, 27 Sep 2022 07:55:03 +0000 (07:55 +0000)]
dt-bindings: crypto: rockchip: add new compatible

Since driver support new compatible, we need to update the driver bindings.

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agodt-bindings: crypto: convert rockchip-crypto to YAML
Corentin Labbe [Tue, 27 Sep 2022 07:55:02 +0000 (07:55 +0000)]
dt-bindings: crypto: convert rockchip-crypto to YAML

Convert rockchip-crypto to YAML.

Reviewed-by: John Keeping <john@metanate.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - use the rk_crypto_info given as parameter
Corentin Labbe [Tue, 27 Sep 2022 07:55:01 +0000 (07:55 +0000)]
crypto: rockchip - use the rk_crypto_info given as parameter

Instead of using the crypto_info from TFM ctx, use the one given as parameter.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - use a rk_crypto_info variable instead of lot of indirection
Corentin Labbe [Tue, 27 Sep 2022 07:55:00 +0000 (07:55 +0000)]
crypto: rockchip - use a rk_crypto_info variable instead of lot of indirection

Instead of using lot of ctx->dev->xx indirections, use an intermediate
variable for rk_crypto_info.
This will help later, when 2 different rk_crypto_info would be used.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - rework rk_handle_req function
Corentin Labbe [Tue, 27 Sep 2022 07:54:59 +0000 (07:54 +0000)]
crypto: rockchip - rework rk_handle_req function

This patch rework the rk_handle_req(), simply removing the
rk_crypto_info parameter.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - rename ablk functions to cipher
Corentin Labbe [Tue, 27 Sep 2022 07:54:58 +0000 (07:54 +0000)]
crypto: rockchip - rename ablk functions to cipher

Some functions have still ablk in their name even if there are
not handling ablk_cipher anymore.
So let's rename them.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - add support for rk3328
Corentin Labbe [Tue, 27 Sep 2022 07:54:57 +0000 (07:54 +0000)]
crypto: rockchip - add support for rk3328

The rk3328 could be used as-is by the rockchip driver.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - fix style issue
Corentin Labbe [Tue, 27 Sep 2022 07:54:56 +0000 (07:54 +0000)]
crypto: rockchip - fix style issue

This patch fixes some warning reported by checkpatch

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - use read_poll_timeout
Corentin Labbe [Tue, 27 Sep 2022 07:54:55 +0000 (07:54 +0000)]
crypto: rockchip - use read_poll_timeout

Use read_poll_timeout instead of open coding it.
In the same time, fix indentation of related comment.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - add myself as maintainer
Corentin Labbe [Tue, 27 Sep 2022 07:54:54 +0000 (07:54 +0000)]
crypto: rockchip - add myself as maintainer

Nobody is set as maintainer of rockchip crypto, I propose to do it as I
have already reworked lot of this code.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - use clk_bulk to simplify clock management
Corentin Labbe [Tue, 27 Sep 2022 07:54:53 +0000 (07:54 +0000)]
crypto: rockchip - use clk_bulk to simplify clock management

rk3328 does not have the same clock names than rk3288, instead of using a complex
clock management, let's use clk_bulk to simplify their handling.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - handle reset also in PM
Corentin Labbe [Tue, 27 Sep 2022 07:54:52 +0000 (07:54 +0000)]
crypto: rockchip - handle reset also in PM

reset could be handled by PM functions.
We keep the initial reset pulse to be sure the hw is a know device state
after probe.

Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - introduce PM
Corentin Labbe [Tue, 27 Sep 2022 07:54:51 +0000 (07:54 +0000)]
crypto: rockchip - introduce PM

Add runtime PM support for rockchip crypto.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - add debugfs
Corentin Labbe [Tue, 27 Sep 2022 07:54:50 +0000 (07:54 +0000)]
crypto: rockchip - add debugfs

This patch enable to access usage stats for each algorithm.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - rewrite type
Corentin Labbe [Tue, 27 Sep 2022 07:54:49 +0000 (07:54 +0000)]
crypto: rockchip - rewrite type

Instead of using a custom type for classify algorithms, let's just use
already defined ones.
And let's made a bit more verbose about what is registered.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - rework by using crypto_engine
Corentin Labbe [Tue, 27 Sep 2022 07:54:48 +0000 (07:54 +0000)]
crypto: rockchip - rework by using crypto_engine

Instead of doing manual queue management, let's use the crypto/engine
for that.
In the same time, rework the requests handling to be easier to
understand (and fix all bugs related to them).

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - remove non-aligned handling
Corentin Labbe [Tue, 27 Sep 2022 07:54:47 +0000 (07:54 +0000)]
crypto: rockchip - remove non-aligned handling

Now driver have fallback for un-aligned cases, remove all code handling
those cases.

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - better handle cipher key
Corentin Labbe [Tue, 27 Sep 2022 07:54:46 +0000 (07:54 +0000)]
crypto: rockchip - better handle cipher key

The key should not be set in hardware too much in advance, this will
fail it 2 TFM with different keys generate alternative requests.
The key should be stored and used just before doing cipher operations.

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - add fallback for ahash
Corentin Labbe [Tue, 27 Sep 2022 07:54:45 +0000 (07:54 +0000)]
crypto: rockchip - add fallback for ahash

Adds a fallback for all case hardware cannot handle.

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - add fallback for cipher
Corentin Labbe [Tue, 27 Sep 2022 07:54:44 +0000 (07:54 +0000)]
crypto: rockchip - add fallback for cipher

The hardware does not handle 0 size length request, let's add a
fallback.
Furthermore fallback will be used for all unaligned case the hardware
cannot handle.

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - do not store mode globally
Corentin Labbe [Tue, 27 Sep 2022 07:54:43 +0000 (07:54 +0000)]
crypto: rockchip - do not store mode globally

Storing the mode globally does not work if 2 requests are handled in the
same time.
We should store it in a request context.

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - fix privete/private typo
Corentin Labbe [Tue, 27 Sep 2022 07:54:42 +0000 (07:54 +0000)]
crypto: rockchip - fix privete/private typo

This fix a simple typo on private word.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - do not do custom power management
Corentin Labbe [Tue, 27 Sep 2022 07:54:41 +0000 (07:54 +0000)]
crypto: rockchip - do not do custom power management

The clock enable/disable at tfm init/exit is fragile,
if 2 tfm are init in the same time and one is removed just after,
it will leave the hardware uncloked even if a user remains.

Instead simply enable clocks at probe time.
We will do PM later.

Fixes: ce0183cb6464b ("crypto: rockchip - switch to skcipher API")
Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - do not use uninitialized variable
Corentin Labbe [Tue, 27 Sep 2022 07:54:40 +0000 (07:54 +0000)]
crypto: rockchip - do not use uninitialized variable

crypto_info->dev is not yet set, so use pdev->dev instead.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: rockchip - use dev_err for error message about interrupt
Corentin Labbe [Tue, 27 Sep 2022 07:54:39 +0000 (07:54 +0000)]
crypto: rockchip - use dev_err for error message about interrupt

Interrupt is mandatory so the message should be printed as error.

Reviewed-by: John Keeping <john@metanate.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agohwrng: stm32 - fix read of the last word
Tomas Marek [Wed, 12 Oct 2022 16:09:24 +0000 (18:09 +0200)]
hwrng: stm32 - fix read of the last word

The stm32_rng_read() function samples TRNG by 4 bytes until at
least 5 bytes are free in the input buffer. The last four bytes
are never read. For example, 60 bytes are returned in case the
input buffer size is 64 bytes.

Read until at least 4 bytes are free in the input buffer. Fill
the buffer entirely in case the buffer size is divisible by 4.

Cc: Oleg Karfich <oleg.karfich@wago.com>
Signed-off-by: Tomas Marek <tomas.marek@elrest.cz>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agohwrng: stm32 - fix number of returned bytes on read
Tomas Marek [Wed, 12 Oct 2022 16:09:23 +0000 (18:09 +0200)]
hwrng: stm32 - fix number of returned bytes on read

The stm32_rng_read() function uses `retval` variable as a counter of
generated random bytes. However, the same variable is used to store
a result of the polling function in case the driver is waiting until
the TRNG is ready. The TRNG generates random numbers by 16B. One
loop read 4B. So, the function calls the polling every 16B, i.e.
every 4th loop. The `retval` counter is reset on poll call and only
number of bytes read after the last poll call is returned to the
caller. The remaining sampled random bytes (for example 48 out of
64 in case 64 bytes are read) are not used.

Use different variable to store the polling function result and
do not overwrite `retval` counter.

Cc: Oleg Karfich <oleg.karfich@wago.com>
Signed-off-by: Tomas Marek <tomas.marek@elrest.cz>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: talitos - Use the defined variable to clean code
jianchunfu [Sun, 9 Oct 2022 09:52:54 +0000 (17:52 +0800)]
crypto: talitos - Use the defined variable to clean code

Use the defined variable "dev" to make the code cleaner.

Signed-off-by: jianchunfu <jianchunfu@cmss.chinamobile.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agohwrng: mtk - add mt7986 support
Mingming.Su [Sat, 8 Oct 2022 16:45:53 +0000 (18:45 +0200)]
hwrng: mtk - add mt7986 support

1. Add trng compatible name for MT7986
2. Fix mtk_rng_wait_ready() function

Signed-off-by: Mingming.Su <Mingming.Su@mediatek.com>
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: n2 - add missing hash statesize
Corentin Labbe [Thu, 6 Oct 2022 04:34:19 +0000 (04:34 +0000)]
crypto: n2 - add missing hash statesize

Add missing statesize to hash templates.
This is mandatory otherwise no algorithms can be registered as the core
requires statesize to be set.

CC: stable@kernel.org # 4.3+
Reported-by: Rolf Eike Beer <eike-kernel@sf-tec.de>
Tested-by: Rolf Eike Beer <eike-kernel@sf-tec.de>
Fixes: 0a625fd2abaa ("crypto: n2 - Add Niagara2 crypto driver")
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: tcrypt - fix return value for multiple subtests
Robert Elliott [Fri, 30 Sep 2022 21:40:14 +0000 (16:40 -0500)]
crypto: tcrypt - fix return value for multiple subtests

When a test mode invokes multiple tests (e.g., mode 0 invokes modes
1 through 199, and mode 3 tests three block cipher modes with des),
don't keep accumulating the return values with ret += tcrypt_test(),
which results in a bogus value if more than one report a nonzero
value (e.g., two reporting -2 (-ENOENT) end up reporting -4 (-EINTR)).
Instead, keep track of the minimum return value reported by any
subtest.

Fixes: 4e033a6bc70f ("crypto: tcrypt - Do not exit on success in fips mode")
Signed-off-by: Robert Elliott <elliott@hpe.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: ixp4xx - Fix sparse warnings
Herbert Xu [Fri, 30 Sep 2022 06:09:34 +0000 (14:09 +0800)]
crypto: ixp4xx - Fix sparse warnings

This fixes a number of trivial sparse warnings in ixp4xx.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Acked-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Tested-by: Corentin Labbe <clabbe.montjoie@gmail.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agohwrng: npcm - Add NPCM8XX support
Tomer Maimon [Thu, 29 Sep 2022 13:31:11 +0000 (16:31 +0300)]
hwrng: npcm - Add NPCM8XX support

Adding RNG NPCM8XX support to NPCM RNG driver.
RNG NPCM8XX uses a different clock prescaler.

As part of adding NPCM8XX support:
- Add NPCM8XX specific compatible string.
- Add data to handle architecture specific clock prescaler.

Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agodt-bindings: rng: nuvoton,npcm-rng: Add npcm845 compatible string
Tomer Maimon [Thu, 29 Sep 2022 13:31:10 +0000 (16:31 +0300)]
dt-bindings: rng: nuvoton,npcm-rng: Add npcm845 compatible string

Add a compatible string for Nuvoton BMC NPCM845 RNG.

Signed-off-by: Tomer Maimon <tmaimon77@gmail.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: stm32 - Fix spelling mistake "wite" -> "write"
Colin Ian King [Wed, 28 Sep 2022 22:24:43 +0000 (23:24 +0100)]
crypto: stm32 - Fix spelling mistake "wite" -> "write"

There are a couple of spelling mistakes in dev_err messages. Fix them.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Acked-by: nicolas.toromanoff@foss.st.com
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: ccp - Add support for TEE for PCI ID 0x14CA
Mario Limonciello [Wed, 28 Sep 2022 18:45:05 +0000 (13:45 -0500)]
crypto: ccp - Add support for TEE for PCI ID 0x14CA

SoCs containing 0x14CA are present both in datacenter parts that
support SEV as well as client parts that support TEE.

Cc: stable@vger.kernel.org # 5.15+
Tested-by: Rijo-john Thomas <Rijo-john.Thomas@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Acked-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: nitrox - avoid double free on error path in nitrox_sriov_init()
Natalia Petrova [Wed, 28 Sep 2022 10:25:05 +0000 (13:25 +0300)]
crypto: nitrox - avoid double free on error path in nitrox_sriov_init()

If alloc_workqueue() fails in nitrox_mbox_init() it deallocates
ndev->iov.vfdev and returns error code, but then nitrox_sriov_init()
calls nitrox_sriov_cleanup() where ndev->iov.vfdev is deallocated
again.

Fix this by nulling ndev->iov.vfdev after the first deallocation.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Fixes: 9e5de3e06e54 ("crypto: cavium/nitrox - Add mailbox...")
Signed-off-by: Natalia Petrova <n.petrova@fintech.ru>
Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: ccp - Remove unused struct ccp_crypto_cpu
Yuan Can [Tue, 27 Sep 2022 13:39:55 +0000 (13:39 +0000)]
crypto: ccp - Remove unused struct ccp_crypto_cpu

After commit bc3854476f36("crypto: ccp - Use a single queue for proper ordering
of tfm requests"), no one use struct ccp_crypto_cpu, so remove it.

Signed-off-by: Yuan Can <yuancan@huawei.com>
Acked-by: John Allen <john.allen@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: sun8i-ss - use dma_addr instead u32
Corentin Labbe [Tue, 27 Sep 2022 08:55:55 +0000 (08:55 +0000)]
crypto: sun8i-ss - use dma_addr instead u32

The DMA address need to be stored in a dma_addr_t

Fixes: 359e893e8af4 ("crypto: sun8i-ss - rework handling of IV")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Corentin Labbe <clabbe@baylibre.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: inside-secure - Add MaxLinear platform
Peter Harliman Liem [Tue, 27 Sep 2022 03:10:10 +0000 (11:10 +0800)]
crypto: inside-secure - Add MaxLinear platform

This is to add MaxLinear platform into compatible id.
Firmware endianness option is added since MaxLinear
firmware is in little endian format.

Signed-off-by: Peter Harliman Liem <pliem@maxlinear.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: inside-secure - Add fw_little_endian option
Peter Harliman Liem [Tue, 27 Sep 2022 03:10:09 +0000 (11:10 +0800)]
crypto: inside-secure - Add fw_little_endian option

This is to add fw_little_endian option, which can
be used for platform which firmware is using little-endian
(instead of big-endian).

Signed-off-by: Peter Harliman Liem <pliem@maxlinear.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: inside-secure - Expand soc data structure
Peter Harliman Liem [Tue, 27 Sep 2022 03:10:08 +0000 (11:10 +0800)]
crypto: inside-secure - Expand soc data structure

Currently platform data is assigned directly to
version string(instead of struct). To make it more
scalable, we move it to use data struct instead.
This allows customization for individual platforms other
than version string.

Signed-off-by: Peter Harliman Liem <pliem@maxlinear.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: talitos - Replace zero-length arrays with DECLARE_FLEX_ARRAY() helper
Gustavo A. R. Silva [Mon, 26 Sep 2022 21:45:45 +0000 (16:45 -0500)]
crypto: talitos - Replace zero-length arrays with DECLARE_FLEX_ARRAY() helper

Zero-length arrays are deprecated and we are moving towards adopting
C99 flexible-array members, instead. So, replace zero-length arrays
declarations in anonymous union with the new DECLARE_FLEX_ARRAY()
helper macro.

This helper allows for flexible-array members in unions.

Link: https://github.com/KSPP/linux/issues/193
Link: https://github.com/KSPP/linux/issues/216
Link: https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: ccp - Add __init/__exit annotations to module init/exit funcs
ruanjinjie [Mon, 26 Sep 2022 09:27:11 +0000 (17:27 +0800)]
crypto: ccp - Add __init/__exit annotations to module init/exit funcs

Add missing __init/__exit annotations to module init/exit funcs

Signed-off-by: ruanjinjie <ruanjinjie@huawei.com>
Acked-by: John Allen <john.allen@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: scatterwalk - remove duplicate function declarations
Tianjia Zhang [Mon, 26 Sep 2022 09:14:40 +0000 (17:14 +0800)]
crypto: scatterwalk - remove duplicate function declarations

scatterwalk_map() is an inline function already defined in the
header file, it is necessary to delete the re-declaration at the
same location, which was left out in the header file by an
earlier modification.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: ccm - use local variables instead of indirect references
Tianjia Zhang [Mon, 26 Sep 2022 09:14:21 +0000 (17:14 +0800)]
crypto: ccm - use local variables instead of indirect references

The variable odata has been introduced into the function scope as
a variable and should be used directly.

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/qm - re-enable communicate interrupt before notifying PF
Weili Qian [Sat, 24 Sep 2022 11:04:31 +0000 (19:04 +0800)]
crypto: hisilicon/qm - re-enable communicate interrupt before notifying PF

After the device is reset, the VF needs to re-enable communication
interrupt before the VF sends restart complete message to the PF.
If the interrupt is re-enabled after the VF notifies the PF, the PF
may fail to send messages to the VF after receiving VF's restart
complete message.

Fixes: 760fe22cf5e9 ("crypto: hisilicon/qm - update reset flow")
Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/sec - enabling clock gating of the address prefetch module
Weili Qian [Sat, 24 Sep 2022 10:34:45 +0000 (18:34 +0800)]
crypto: hisilicon/sec - enabling clock gating of the address prefetch module

Change the value of clock gating register to 0x7fff to enable
clock gating of the address prefetch module. When the device is
idle, the clock is turned off to save power.

Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/qm - fix incorrect parameters usage
Weili Qian [Sat, 24 Sep 2022 10:14:42 +0000 (18:14 +0800)]
crypto: hisilicon/qm - fix incorrect parameters usage

In qm_get_xqc_depth(), parameters low_bits and high_bits save
the values of the corresponding bits. However, the values saved by the
two parameters are opposite. As a result, the values returned to the
callers are incorrect.

Fixes: 129a9f340172 ("crypto: hisilicon/qm - get qp num and depth from hardware registers")
Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/qm - drop unnecessary IS_ENABLE(CONFIG_NUMA) check
Yicong Yang [Sat, 24 Sep 2022 09:34:24 +0000 (17:34 +0800)]
crypto: hisilicon/qm - drop unnecessary IS_ENABLE(CONFIG_NUMA) check

dev_to_node() can handle the case when CONFIG_NUMA is not set, so the
check of CONFIG_NUMA is redundant and can be removed.

Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
Signed-off-by: Weili Qian <qianweili@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: hisilicon/hpre - fix resource leak in remove process
Zhiqi Song [Sat, 24 Sep 2022 07:38:31 +0000 (15:38 +0800)]
crypto: hisilicon/hpre - fix resource leak in remove process

In hpre_remove(), when the disable operation of qm sriov failed,
the following logic should continue to be executed to release the
remaining resources that have been allocated, instead of returning
directly, otherwise there will be resource leakage.

Signed-off-by: Zhiqi Song <songzhiqi1@huawei.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agocrypto: x86/polyval - Fix crashes when keys are not 16-byte aligned
Nathan Huckleberry [Tue, 18 Oct 2022 23:04:12 +0000 (16:04 -0700)]
crypto: x86/polyval - Fix crashes when keys are not 16-byte aligned

crypto_tfm::__crt_ctx is not guaranteed to be 16-byte aligned on x86-64.
This causes crashes due to movaps instructions in clmul_polyval_update.

Add logic to align polyval_tfm_ctx to 16 bytes.

Cc: <stable@vger.kernel.org>
Fixes: 34f7f6c30112 ("crypto: x86/polyval - Add PCLMULQDQ accelerated implementation of POLYVAL")
Reported-by: Bruno Goncalves <bgoncalv@redhat.com>
Signed-off-by: Nathan Huckleberry <nhuck@google.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2 years agoLinux 6.1-rc1
Linus Torvalds [Sun, 16 Oct 2022 22:36:24 +0000 (15:36 -0700)]
Linux 6.1-rc1

2 years agoMerge tag 'random-6.1-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 16 Oct 2022 22:27:07 +0000 (15:27 -0700)]
Merge tag 'random-6.1-rc1-for-linus' of git://git./linux/kernel/git/crng/random

Pull more random number generator updates from Jason Donenfeld:
 "This time with some large scale treewide cleanups.

  The intent of this pull is to clean up the way callers fetch random
  integers. The current rules for doing this right are:

   - If you want a secure or an insecure random u64, use get_random_u64()

   - If you want a secure or an insecure random u32, use get_random_u32()

     The old function prandom_u32() has been deprecated for a while
     now and is just a wrapper around get_random_u32(). Same for
     get_random_int().

   - If you want a secure or an insecure random u16, use get_random_u16()

   - If you want a secure or an insecure random u8, use get_random_u8()

   - If you want secure or insecure random bytes, use get_random_bytes().

     The old function prandom_bytes() has been deprecated for a while
     now and has long been a wrapper around get_random_bytes()

   - If you want a non-uniform random u32, u16, or u8 bounded by a
     certain open interval maximum, use prandom_u32_max()

     I say "non-uniform", because it doesn't do any rejection sampling
     or divisions. Hence, it stays within the prandom_*() namespace, not
     the get_random_*() namespace.

     I'm currently investigating a "uniform" function for 6.2. We'll see
     what comes of that.

  By applying these rules uniformly, we get several benefits:

   - By using prandom_u32_max() with an upper-bound that the compiler
     can prove at compile-time is ≤65536 or ≤256, internally
     get_random_u16() or get_random_u8() is used, which wastes fewer
     batched random bytes, and hence has higher throughput.

   - By using prandom_u32_max() instead of %, when the upper-bound is
     not a constant, division is still avoided, because
     prandom_u32_max() uses a faster multiplication-based trick instead.

   - By using get_random_u16() or get_random_u8() in cases where the
     return value is intended to indeed be a u16 or a u8, we waste fewer
     batched random bytes, and hence have higher throughput.

  This series was originally done by hand while I was on an airplane
  without Internet. Later, Kees and I worked on retroactively figuring
  out what could be done with Coccinelle and what had to be done
  manually, and then we split things up based on that.

  So while this touches a lot of files, the actual amount of code that's
  hand fiddled is comfortably small"

* tag 'random-6.1-rc1-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/crng/random:
  prandom: remove unused functions
  treewide: use get_random_bytes() when possible
  treewide: use get_random_u32() when possible
  treewide: use get_random_{u8,u16}() when possible, part 2
  treewide: use get_random_{u8,u16}() when possible, part 1
  treewide: use prandom_u32_max() when possible, part 2
  treewide: use prandom_u32_max() when possible, part 1

2 years agoMerge tag 'perf-tools-for-v6.1-2-2022-10-16' of git://git.kernel.org/pub/scm/linux...
Linus Torvalds [Sun, 16 Oct 2022 22:14:29 +0000 (15:14 -0700)]
Merge tag 'perf-tools-for-v6.1-2-2022-10-16' of git://git./linux/kernel/git/acme/linux

Pull more perf tools updates from Arnaldo Carvalho de Melo:

 - Use BPF CO-RE (Compile Once, Run Everywhere) to support old kernels
   when using bperf (perf BPF based counters) with cgroups.

 - Support HiSilicon PCIe Performance Monitoring Unit (PMU), that
   monitors bandwidth, latency, bus utilization and buffer occupancy.

   Documented in Documentation/admin-guide/perf/hisi-pcie-pmu.rst.

 - User space tasks can migrate between CPUs, so when tracing selected
   CPUs, system-wide sideband is still needed, fix it in the setup of
   Intel PT on hybrid systems.

 - Fix metricgroups title message in 'perf list', it should state that
   the metrics groups are to be used with the '-M' option, not '-e'.

 - Sync the msr-index.h copy with the kernel sources, adding support for
   using "AMD64_TSC_RATIO" in filter expressions in 'perf trace' as well
   as decoding it when printing the MSR tracepoint arguments.

 - Fix program header size and alignment when generating a JIT ELF in
   'perf inject'.

 - Add multiple new Intel PT 'perf test' entries, including a jitdump
   one.

 - Fix the 'perf test' entries for 'perf stat' CSV and JSON output when
   running on PowerPC due to an invalid topology number in that arch.

 - Fix the 'perf test' for arm_coresight failures on the ARM Juno
   system.

 - Fix the 'perf test' attr entry for PERF_FORMAT_LOST, adding this
   option to the or expression expected in the intercepted
   perf_event_open() syscall.

 - Add missing condition flags ('hs', 'lo', 'vc', 'vs') for arm64 in the
   'perf annotate' asm parser.

 - Fix 'perf mem record -C' option processing, it was being chopped up
   when preparing the underlying 'perf record -e mem-events' and thus
   being ignored, requiring using '-- -C CPUs' as a workaround.

 - Improvements and tidy ups for 'perf test' shell infra.

 - Fix Intel PT information printing segfault in uClibc, where a NULL
   format was being passed to fprintf.

* tag 'perf-tools-for-v6.1-2-2022-10-16' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (23 commits)
  tools arch x86: Sync the msr-index.h copy with the kernel sources
  perf auxtrace arm64: Add support for parsing HiSilicon PCIe Trace packet
  perf auxtrace arm64: Add support for HiSilicon PCIe Tune and Trace device driver
  perf auxtrace arm: Refactor event list iteration in auxtrace_record__init()
  perf tests stat+json_output: Include sanity check for topology
  perf tests stat+csv_output: Include sanity check for topology
  perf intel-pt: Fix system_wide dummy event for hybrid
  perf intel-pt: Fix segfault in intel_pt_print_info() with uClibc
  perf test: Fix attr tests for PERF_FORMAT_LOST
  perf test: test_intel_pt.sh: Add 9 tests
  perf inject: Fix GEN_ELF_TEXT_OFFSET for jit
  perf test: test_intel_pt.sh: Add jitdump test
  perf test: test_intel_pt.sh: Tidy some alignment
  perf test: test_intel_pt.sh: Print a message when skipping kernel tracing
  perf test: test_intel_pt.sh: Tidy some perf record options
  perf test: test_intel_pt.sh: Fix return checking again
  perf: Skip and warn on unknown format 'configN' attrs
  perf list: Fix metricgroups title message
  perf mem: Fix -C option behavior for perf mem record
  perf annotate: Add missing condition flags for arm64
  ...

2 years agoMerge tag 'kbuild-fixes-v6.1' of git://git.kernel.org/pub/scm/linux/kernel/git/masahi...
Linus Torvalds [Sun, 16 Oct 2022 18:12:22 +0000 (11:12 -0700)]
Merge tag 'kbuild-fixes-v6.1' of git://git./linux/kernel/git/masahiroy/linux-kbuild

Pull Kbuild fixes from Masahiro Yamada:

 - Fix CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y compile error for the
   combination of Clang >= 14 and GAS <= 2.35.

 - Drop vmlinux.bz2 from the rpm package as it just annoyingly increased
   the package size.

 - Fix modpost error under build environments using musl.

 - Make *.ll files keep value names for easier debugging

 - Fix single directory build

 - Prevent RISC-V from selecting the broken DWARF5 support when Clang
   and GAS are used together.

* tag 'kbuild-fixes-v6.1' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
  lib/Kconfig.debug: Add check for non-constant .{s,u}leb128 support to DWARF5
  kbuild: fix single directory build
  kbuild: add -fno-discard-value-names to cmd_cc_ll_c
  scripts/clang-tools: Convert clang-tidy args to list
  modpost: put modpost options before argument
  kbuild: Stop including vmlinux.bz2 in the rpm's
  Kconfig.debug: add toolchain checks for DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT
  Kconfig.debug: simplify the dependency of DEBUG_INFO_DWARF4/5

2 years agoMerge tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Linus Torvalds [Sun, 16 Oct 2022 18:08:19 +0000 (11:08 -0700)]
Merge tag 'clk-for-linus' of git://git./linux/kernel/git/clk/linux

Pull more clk updates from Stephen Boyd:
 "This is the final part of the clk patches for this merge window.

  The clk rate range series needed another week to fully bake. Maxime
  fixed the bug that broke clk notifiers and prevented this from being
  included in the first pull request. He also added a unit test on top
  to make sure it doesn't break so easily again. The majority of the
  series fixes up how the clk_set_rate_*() APIs work, particularly
  around when the rate constraints are dropped and how they move around
  when reparenting clks. Overall it's a much needed improvement to the
  clk rate range APIs that used to be pretty broken if you looked
  sideways.

  Beyond the core changes there are a few driver fixes for a compilation
  issue or improper data causing clks to fail to register or have the
  wrong parents. These are good to get in before the first -rc so that
  the system actually boots on the affected devices"

* tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux: (31 commits)
  clk: tegra: Fix Tegra PWM parent clock
  clk: at91: fix the build with binutils 2.27
  clk: qcom: gcc-msm8660: Drop hardcoded fixed board clocks
  clk: mediatek: clk-mux: Add .determine_rate() callback
  clk: tests: Add tests for notifiers
  clk: Update req_rate on __clk_recalc_rates()
  clk: tests: Add missing test case for ranges
  clk: qcom: clk-rcg2: Take clock boundaries into consideration for gfx3d
  clk: Introduce the clk_hw_get_rate_range function
  clk: Zero the clk_rate_request structure
  clk: Stop forwarding clk_rate_requests to the parent
  clk: Constify clk_has_parent()
  clk: Introduce clk_core_has_parent()
  clk: Switch from __clk_determine_rate to clk_core_round_rate_nolock
  clk: Add our request boundaries in clk_core_init_rate_req
  clk: Introduce clk_hw_init_rate_request()
  clk: Move clk_core_init_rate_req() from clk_core_round_rate_nolock() to its caller
  clk: Change clk_core_init_rate_req prototype
  clk: Set req_rate on reparenting
  clk: Take into account uncached clocks in clk_set_rate_range()
  ...

2 years agoMerge tag '6.1-rc-smb3-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Sun, 16 Oct 2022 18:01:40 +0000 (11:01 -0700)]
Merge tag '6.1-rc-smb3-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6

Pull more cifs updates from Steve French:

 - fix a regression in guest mounts to old servers

 - improvements to directory leasing (caching directory entries safely
   beyond the root directory)

 - symlink improvement (reducing roundtrips needed to process symlinks)

 - an lseek fix (to problem where some dir entries could be skipped)

 - improved ioctl for returning more detailed information on directory
   change notifications

 - clarify multichannel interface query warning

 - cleanup fix (for better aligning buffers using ALIGN and round_up)

 - a compounding fix

 - fix some uninitialized variable bugs found by Coverity and the kernel
   test robot

* tag '6.1-rc-smb3-client-fixes-part2' of git://git.samba.org/sfrench/cifs-2.6:
  smb3: improve SMB3 change notification support
  cifs: lease key is uninitialized in two additional functions when smb1
  cifs: lease key is uninitialized in smb1 paths
  smb3: must initialize two ACL struct fields to zero
  cifs: fix double-fault crash during ntlmssp
  cifs: fix static checker warning
  cifs: use ALIGN() and round_up() macros
  cifs: find and use the dentry for cached non-root directories also
  cifs: enable caching of directories for which a lease is held
  cifs: prevent copying past input buffer boundaries
  cifs: fix uninitialised var in smb2_compound_op()
  cifs: improve symlink handling for smb2+
  smb3: clarify multichannel warning
  cifs: fix regression in very old smb1 mounts
  cifs: fix skipping to incorrect offset in emit_cached_dirents

2 years agoRevert "cpumask: fix checking valid cpu range".
Tetsuo Handa [Sat, 15 Oct 2022 15:53:51 +0000 (00:53 +0900)]
Revert "cpumask: fix checking valid cpu range".

This reverts commit 78e5a3399421 ("cpumask: fix checking valid cpu range").

syzbot is hitting WARN_ON_ONCE(cpu >= nr_cpumask_bits) warning at
cpu_max_bits_warn() [1], for commit 78e5a3399421 ("cpumask: fix checking
valid cpu range") is broken.  Obviously that patch hits WARN_ON_ONCE()
when e.g.  reading /proc/cpuinfo because passing "cpu + 1" instead of
"cpu" will trivially hit cpu == nr_cpumask_bits condition.

Although syzbot found this problem in linux-next.git on 2022/09/27 [2],
this problem was not fixed immediately.  As a result, that patch was
sent to linux.git before the patch author recognizes this problem, and
syzbot started failing to test changes in linux.git since 2022/10/10
[3].

Andrew Jones proposed a fix for x86 and riscv architectures [4].  But
[2] and [5] indicate that affected locations are not limited to arch
code.  More delay before we find and fix affected locations, less tested
kernel (and more difficult to bisect and fix) before release.

We should have inspected and fixed basically all cpumask users before
applying that patch.  We should not crash kernels in order to ask
existing cpumask users to update their code, even if limited to
CONFIG_DEBUG_PER_CPU_MAPS=y case.

Link: https://syzkaller.appspot.com/bug?extid=d0fd2bf0dd6da72496dd
Link: https://syzkaller.appspot.com/bug?extid=21da700f3c9f0bc40150
Link: https://syzkaller.appspot.com/bug?extid=51a652e2d24d53e75734
Link: https://lkml.kernel.org/r/20221014155845.1986223-1-ajones@ventanamicro.com
Link: https://syzkaller.appspot.com/bug?extid=4d46c43d81c3bd155060
Reported-by: Andrew Jones <ajones@ventanamicro.com>
Reported-by: syzbot+d0fd2bf0dd6da72496dd@syzkaller.appspotmail.com
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: Yury Norov <yury.norov@gmail.com>
Cc: Borislav Petkov <bp@alien8.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 years agolib/Kconfig.debug: Add check for non-constant .{s,u}leb128 support to DWARF5
Nathan Chancellor [Fri, 14 Oct 2022 20:42:11 +0000 (13:42 -0700)]
lib/Kconfig.debug: Add check for non-constant .{s,u}leb128 support to DWARF5

When building with a RISC-V kernel with DWARF5 debug info using clang
and the GNU assembler, several instances of the following error appear:

  /tmp/vgettimeofday-48aa35.s:2963: Error: non-constant .uleb128 is not supported

Dumping the .s file reveals these .uleb128 directives come from
.debug_loc and .debug_ranges:

  .Ldebug_loc0:
          .byte   4                               # DW_LLE_offset_pair
          .uleb128 .Lfunc_begin0-.Lfunc_begin0    #   starting offset
          .uleb128 .Ltmp1-.Lfunc_begin0           #   ending offset
          .byte   1                               # Loc expr size
          .byte   90                              # DW_OP_reg10
          .byte   0                               # DW_LLE_end_of_list

  .Ldebug_ranges0:
          .byte   4                               # DW_RLE_offset_pair
          .uleb128 .Ltmp6-.Lfunc_begin0           #   starting offset
          .uleb128 .Ltmp27-.Lfunc_begin0          #   ending offset
          .byte   4                               # DW_RLE_offset_pair
          .uleb128 .Ltmp28-.Lfunc_begin0          #   starting offset
          .uleb128 .Ltmp30-.Lfunc_begin0          #   ending offset
          .byte   0                               # DW_RLE_end_of_list

There is an outstanding binutils issue to support a non-constant operand
to .sleb128 and .uleb128 in GAS for RISC-V but there does not appear to
be any movement on it, due to concerns over how it would work with
linker relaxation.

To avoid these build errors, prevent DWARF5 from being selected when
using clang and an assembler that does not have support for these symbol
deltas, which can be easily checked in Kconfig with as-instr plus the
small test program from the dwz test suite from the binutils issue.

Link: https://sourceware.org/bugzilla/show_bug.cgi?id=27215
Link: https://github.com/ClangBuiltLinux/linux/issues/1719
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
2 years agokbuild: fix single directory build
Masahiro Yamada [Fri, 14 Oct 2022 20:18:11 +0000 (05:18 +0900)]
kbuild: fix single directory build

Commit f110e5a250e3 ("kbuild: refactor single builds of *.ko") was wrong.

KBUILD_MODULES _is_ needed for single builds.

Otherwise, "make foo/bar/baz/" does not build module objects at all.

Fixes: f110e5a250e3 ("kbuild: refactor single builds of *.ko")
Reported-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Tested-by: David Sterba <dsterba@suse.com>
2 years agoMerge tag 'slab-for-6.1-rc1-hotfix' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 16 Oct 2022 00:05:07 +0000 (17:05 -0700)]
Merge tag 'slab-for-6.1-rc1-hotfix' of git://git./linux/kernel/git/vbabka/slab

Pull slab hotfix from Vlastimil Babka:
 "A single fix for the common-kmalloc series, for warnings on mips and
  sparc64 reported by Guenter Roeck"

* tag 'slab-for-6.1-rc1-hotfix' of git://git.kernel.org/pub/scm/linux/kernel/git/vbabka/slab:
  mm/slab: use kmalloc_node() for off slab freelist_idx_t array allocation

2 years agoMerge tag 'for-linus' of https://github.com/openrisc/linux
Linus Torvalds [Sat, 15 Oct 2022 23:47:33 +0000 (16:47 -0700)]
Merge tag 'for-linus' of https://github.com/openrisc/linux

Pull OpenRISC updates from Stafford Horne:
 "I have relocated to London so not much work from me while I get
  settled.

  Still, OpenRISC picked up two patches in this window:

   - Fix for kernel page table walking from Jann Horn

   - MAINTAINER entry cleanup from Palmer Dabbelt"

* tag 'for-linus' of https://github.com/openrisc/linux:
  MAINTAINERS: git://github -> https://github.com for openrisc
  openrisc: Fix pagewalk usage in arch_dma_{clear, set}_uncached

2 years agoMerge tag 'pci-v6.1-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci
Linus Torvalds [Sat, 15 Oct 2022 23:36:38 +0000 (16:36 -0700)]
Merge tag 'pci-v6.1-fixes-1' of git://git./linux/kernel/git/helgaas/pci

Pull pci fix from Bjorn Helgaas:
 "Revert the attempt to distribute spare resources to unconfigured
  hotplug bridges at boot time.

  This fixed some dock hot-add scenarios, but Jonathan Cameron reported
  that it broke a topology with a multi-function device where one
  function was a Switch Upstream Port and the other was an Endpoint"

* tag 'pci-v6.1-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
  Revert "PCI: Distribute available resources for root buses, too"

2 years agomm/slab: use kmalloc_node() for off slab freelist_idx_t array allocation
Hyeonggon Yoo [Sat, 15 Oct 2022 04:34:29 +0000 (13:34 +0900)]
mm/slab: use kmalloc_node() for off slab freelist_idx_t array allocation

After commit d6a71648dbc0 ("mm/slab: kmalloc: pass requests larger than
order-1 page to page allocator"), SLAB passes large ( > PAGE_SIZE * 2)
requests to buddy like SLUB does.

SLAB has been using kmalloc caches to allocate freelist_idx_t array for
off slab caches. But after the commit, freelist_size can be bigger than
KMALLOC_MAX_CACHE_SIZE.

Instead of using pointer to kmalloc cache, use kmalloc_node() and only
check if the kmalloc cache is off slab during calculate_slab_order().
If freelist_size > KMALLOC_MAX_CACHE_SIZE, no looping condition happens
as it allocates freelist_idx_t array directly from buddy.

Link: https://lore.kernel.org/all/20221014205818.GA1428667@roeck-us.net/
Reported-and-tested-by: Guenter Roeck <linux@roeck-us.net>
Fixes: d6a71648dbc0 ("mm/slab: kmalloc: pass requests larger than order-1 page to page allocator")
Signed-off-by: Hyeonggon Yoo <42.hyeyoo@gmail.com>
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
2 years agoMAINTAINERS: git://github -> https://github.com for openrisc
Palmer Dabbelt [Thu, 13 Oct 2022 21:46:37 +0000 (14:46 -0700)]
MAINTAINERS: git://github -> https://github.com for openrisc

Github deprecated the git:// links about a year ago, so let's move to
the https:// URLs instead.

Reported-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://github.blog/2021-09-01-improving-git-protocol-security-github/
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Stafford Horne <shorne@gmail.com>