* [PATCH v4 1/2] crypto: padlock-sha - Disable for Zhaoxin processor
2026-03-13 8:01 [PATCH v4 0/2] lib/crypto: x86/sha: Add PHE Extensions support AlanSong-oc
@ 2026-03-13 8:01 ` AlanSong-oc
2026-03-14 18:40 ` Eric Biggers
2026-03-13 8:01 ` [PATCH v4 2/2] lib/crypto: x86/sha256: PHE Extensions optimized SHA256 transform function AlanSong-oc
1 sibling, 1 reply; 5+ messages in thread
From: AlanSong-oc @ 2026-03-13 8:01 UTC (permalink / raw)
To: herbert, davem, ebiggers, Jason, ardb, linux-crypto, linux-kernel,
x86
Cc: CobeChen, TonyWWang-oc, YunShen, GeorgeXue, LeoLiu, HansHu,
AlanSong-oc, stable
For Zhaoxin processors, the XSHA1 instruction requires the total memory
allocated at %rdi register must be 32 bytes, while the XSHA1 and
XSHA256 instruction doesn't perform any operation when %ecx is zero.
Due to these requirements, the current padlock-sha driver does not work
correctly with Zhaoxin processors. It cannot pass the self-tests and
therefore does not activate the driver on Zhaoxin processors. This issue
has been reported in Debian [1]. The self-tests fail with the
following messages [2]:
alg: shash: sha1-padlock-nano test failed (wrong result) on test vector 0, cfg="init+update+final aligned buffer"
alg: self-tests for sha1 using sha1-padlock-nano failed (rc=-22)
------------[ cut here ]------------
alg: shash: sha256-padlock-nano test failed (wrong result) on test vector 0, cfg="init+update+final aligned buffer"
alg: self-tests for sha256 using sha256-padlock-nano failed (rc=-22)
------------[ cut here ]------------
Disable the padlock-sha driver on Zhaoxin processors with the CPU family
0x07 and newer. Following the suggestion in [3], add support for the PHE
extensions to lib/crypto. Only XSHA256 support for SHA-256 is included,
since SHA-1 has been cryptographically broken, as recommended in [4].
[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1103397
[2] https://linux-hardware.org/?probe=271fabb7a4&log=dmesg
[3] https://lore.kernel.org/linux-crypto/aUI4CGp6kK7mxgEr@gondor.apana.org.au/
[4] https://lore.kernel.org/linux-crypto/20260116071513.12134-1-AlanSong-oc@zhaoxin.com/T/#m49436c4849dd64454b3554c105197ef9c61db23e
Fixes: 63dc06cd12f9 ("crypto: padlock-sha - Use API partial block handling")
Cc: stable@vger.kernel.org
Signed-off-by: AlanSong-oc <AlanSong-oc@zhaoxin.com>
---
drivers/crypto/padlock-sha.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/crypto/padlock-sha.c b/drivers/crypto/padlock-sha.c
index 329f60ad4..9214bbfc8 100644
--- a/drivers/crypto/padlock-sha.c
+++ b/drivers/crypto/padlock-sha.c
@@ -332,6 +332,13 @@ static int __init padlock_init(void)
if (!x86_match_cpu(padlock_sha_ids) || !boot_cpu_has(X86_FEATURE_PHE_EN))
return -ENODEV;
+ /*
+ * Skip family 0x07 and newer used by Zhaoxin processors,
+ * as the driver's self-tests fail on these CPUs.
+ */
+ if (c->x86 >= 0x07)
+ return -ENODEV;
+
/* Register the newly added algorithm module if on *
* VIA Nano processor, or else just do as before */
if (c->x86_model < 0x0f) {
--
2.34.1
^ permalink raw reply related [flat|nested] 5+ messages in thread* [PATCH v4 2/2] lib/crypto: x86/sha256: PHE Extensions optimized SHA256 transform function
2026-03-13 8:01 [PATCH v4 0/2] lib/crypto: x86/sha: Add PHE Extensions support AlanSong-oc
2026-03-13 8:01 ` [PATCH v4 1/2] crypto: padlock-sha - Disable for Zhaoxin processor AlanSong-oc
@ 2026-03-13 8:01 ` AlanSong-oc
2026-03-14 18:50 ` Eric Biggers
1 sibling, 1 reply; 5+ messages in thread
From: AlanSong-oc @ 2026-03-13 8:01 UTC (permalink / raw)
To: herbert, davem, ebiggers, Jason, ardb, linux-crypto, linux-kernel,
x86
Cc: CobeChen, TonyWWang-oc, YunShen, GeorgeXue, LeoLiu, HansHu,
AlanSong-oc
Zhaoxin CPUs have implemented the SHA(Secure Hash Algorithm) as its CPU
instructions by PHE(Padlock Hash Engine) Extensions, including XSHA1,
XSHA256, XSHA384 and XSHA512 instructions. The instruction specification
is available at the following link.
(https://gitee.com/openzhaoxin/zhaoxin_specifications/blob/20260227/ZX_Padlock_Reference.pdf)
With the help of implementation of SHA in hardware instead of software,
can develop applications with higher performance, more security and more
flexibility.
This patch includes the XSHA256 instruction optimized implementation of
SHA-256 transform function.
The table below shows the benchmark results before and after applying
this patch by using CRYPTO_LIB_BENCHMARK on Zhaoxin KX-7000 platform,
highlighting the achieved speedups.
+---------+--------------------------+
| | SHA256 |
+---------+--------+-----------------+
| Len | Before | After |
+---------+--------+-----------------+
| 1* | 2 | 7 (3.50x) |
| 16 | 35 | 119 (3.40x) |
| 64 | 74 | 280 (3.78x) |
| 127 | 99 | 387 (3.91x) |
| 128 | 103 | 427 (4.15x) |
| 200 | 123 | 537 (4.37x) |
| 256 | 128 | 582 (4.55x) |
| 511 | 144 | 679 (4.72x) |
| 512 | 146 | 714 (4.89x) |
| 1024 | 157 | 796 (5.07x) |
| 3173 | 167 | 883 (5.28x) |
| 4096 | 166 | 876 (5.28x) |
| 16384 | 169 | 899 (5.32x) |
+---------+--------+-----------------+
*: The length of each data block to be processed by one complete SHA
sequence.
**: The throughput of processing data blocks, unit is Mb/s.
After applying this patch, the SHA256 KUnit test suite passes on Zhaoxin
platforms. Detailed test logs are shown below.
[ 7.767257] # Subtest: sha256
[ 7.770542] # module: sha256_kunit
[ 7.770544] 1..15
[ 7.777383] ok 1 test_hash_test_vectors
[ 7.788563] ok 2 test_hash_all_lens_up_to_4096
[ 7.806090] ok 3 test_hash_incremental_updates
[ 7.813553] ok 4 test_hash_buffer_overruns
[ 7.822384] ok 5 test_hash_overlaps
[ 7.829388] ok 6 test_hash_alignment_consistency
[ 7.833843] ok 7 test_hash_ctx_zeroization
[ 7.915191] ok 8 test_hash_interrupt_context_1
[ 8.362312] ok 9 test_hash_interrupt_context_2
[ 8.401607] ok 10 test_hmac
[ 8.415458] ok 11 test_sha256_finup_2x
[ 8.419397] ok 12 test_sha256_finup_2x_defaultctx
[ 8.424107] ok 13 test_sha256_finup_2x_hugelen
[ 8.451289] # benchmark_hash: len=1: 7 MB/s
[ 8.465372] # benchmark_hash: len=16: 119 MB/s
[ 8.481760] # benchmark_hash: len=64: 280 MB/s
[ 8.499344] # benchmark_hash: len=127: 387 MB/s
[ 8.515800] # benchmark_hash: len=128: 427 MB/s
[ 8.531970] # benchmark_hash: len=200: 537 MB/s
[ 8.548241] # benchmark_hash: len=256: 582 MB/s
[ 8.564838] # benchmark_hash: len=511: 679 MB/s
[ 8.580872] # benchmark_hash: len=512: 714 MB/s
[ 8.596858] # benchmark_hash: len=1024: 796 MB/s
[ 8.612567] # benchmark_hash: len=3173: 883 MB/s
[ 8.628546] # benchmark_hash: len=4096: 876 MB/s
[ 8.644482] # benchmark_hash: len=16384: 899 MB/s
[ 8.649773] ok 14 benchmark_hash
[ 8.655505] ok 15 benchmark_sha256_finup_2x # SKIP not relevant
[ 8.659065] # sha256: pass:14 fail:0 skip:1 total:15
[ 8.665276] # Totals: pass:14 fail:0 skip:1 total:15
[ 8.670195] ok 7 sha256
Signed-off-by: AlanSong-oc <AlanSong-oc@zhaoxin.com>
---
lib/crypto/x86/sha256.h | 25 +++++++++++++++++++++++++
1 file changed, 25 insertions(+)
diff --git a/lib/crypto/x86/sha256.h b/lib/crypto/x86/sha256.h
index 38e33b22a..5816b8928 100644
--- a/lib/crypto/x86/sha256.h
+++ b/lib/crypto/x86/sha256.h
@@ -31,6 +31,27 @@ DEFINE_X86_SHA256_FN(sha256_blocks_avx, sha256_transform_avx);
DEFINE_X86_SHA256_FN(sha256_blocks_avx2, sha256_transform_rorx);
DEFINE_X86_SHA256_FN(sha256_blocks_ni, sha256_ni_transform);
+#define PHE_ALIGNMENT 16
+static void sha256_blocks_phe(struct sha256_block_state *state,
+ const u8 *data, size_t nblocks)
+{
+ /*
+ * On Zhaoxin processors, XSHA256 requires the %rdi register
+ * in 64-bit mode (or %edi in 32-bit mode) to point to
+ * a 32-byte, 16-byte-aligned buffer.
+ */
+ u8 buf[32 + PHE_ALIGNMENT - 1];
+ u8 *dst = PTR_ALIGN(&buf[0], PHE_ALIGNMENT);
+ size_t padding = -1;
+
+ memcpy(dst, state, SHA256_DIGEST_SIZE);
+ asm volatile(".byte 0xf3,0x0f,0xa6,0xd0" /* REP XSHA256 */
+ : "+a"(padding), "+c"(nblocks), "+S"(data)
+ : "D"(dst)
+ : "memory");
+ memcpy(state, dst, SHA256_DIGEST_SIZE);
+}
+
static void sha256_blocks(struct sha256_block_state *state,
const u8 *data, size_t nblocks)
{
@@ -79,6 +100,10 @@ static void sha256_mod_init_arch(void)
if (boot_cpu_has(X86_FEATURE_SHA_NI)) {
static_call_update(sha256_blocks_x86, sha256_blocks_ni);
static_branch_enable(&have_sha_ni);
+ } else if (IS_ENABLED(CONFIG_CPU_SUP_ZHAOXIN) &&
+ boot_cpu_has(X86_FEATURE_PHE_EN) &&
+ boot_cpu_data.x86 >= 0x07) {
+ static_call_update(sha256_blocks_x86, sha256_blocks_phe);
} else if (cpu_has_xfeatures(XFEATURE_MASK_SSE | XFEATURE_MASK_YMM,
NULL) &&
boot_cpu_has(X86_FEATURE_AVX)) {
--
2.34.1
^ permalink raw reply related [flat|nested] 5+ messages in thread