From nobody Wed Sep 17 14:00:28 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3B186C001B2 for ; Mon, 19 Dec 2022 20:37:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232507AbiLSUhu (ORCPT ); Mon, 19 Dec 2022 15:37:50 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52798 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232394AbiLSUhq (ORCPT ); Mon, 19 Dec 2022 15:37:46 -0500 Received: from mx0a-002e3701.pphosted.com (mx0a-002e3701.pphosted.com [148.163.147.86]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 5AFDAD2FC; Mon, 19 Dec 2022 12:37:44 -0800 (PST) Received: from pps.filterd (m0148663.ppops.net [127.0.0.1]) by mx0a-002e3701.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 2BJKVxbp001379; Mon, 19 Dec 2022 20:37:42 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=hpe.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : mime-version : content-transfer-encoding; s=pps0720; bh=hxDUwTsXgf/F0pCYDtIB+NfpqhGa9UxYzYrva7hoWO4=; b=bdwGlHDmJVGNT9TxaBzhIRCsCpWGGZL9J3HKrwKqG0UgBVE4QKsX5bKKsTGLb805qo/r SU48jqI/CfQ3icUgybiUw96ANFbgN80lonJ3WlOvJ7tjzlE+Yt8XcI+S7ccI5h9aHee5 WwlMMq2/BtpG4FOCbbxs+37klzwgV4oUBVgi5qQnqmTfRWcL4/E8JNvLKrjY/2EiVC+h glYj/o8abVUUVYloNY34Hyv7lZdkV09+GZs0RnG+8ljgv3cfTcYKOPUJd2mP5lpkq0Ls 3iUzsEozcTMHjES7krXLZOI7M4OhGVQzZ+eeUjbl8IUM9W/+wUVPBhlpaqglTgJ4KubG dA== Received: from p1lg14881.it.hpe.com (p1lg14881.it.hpe.com [16.230.97.202]) by mx0a-002e3701.pphosted.com (PPS) with ESMTPS id 3mjx3mrg6r-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 19 Dec 2022 20:37:41 +0000 Received: from p1lg14886.dc01.its.hpecorp.net (unknown [10.119.18.237]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by p1lg14881.it.hpe.com (Postfix) with ESMTPS id 52F46801723; Mon, 19 Dec 2022 20:37:41 +0000 (UTC) Received: from adevxp033-sys.us.rdlabs.hpecorp.net (unknown [16.231.227.36]) by p1lg14886.dc01.its.hpecorp.net (Postfix) with ESMTP id 0A8F0807B0B; Mon, 19 Dec 2022 20:37:41 +0000 (UTC) From: Robert Elliott To: herbert@gondor.apana.org.au, davem@davemloft.net Cc: linux-crypto@vger.kernel.org, linux-kernel@vger.kernel.org, Robert Elliott Subject: [PATCH 1/3] crypto: skcipher - always yield at end of walk Date: Mon, 19 Dec 2022 14:37:31 -0600 Message-Id: <20221219203733.3063192-2-elliott@hpe.com> X-Mailer: git-send-email 2.38.1 In-Reply-To: <20221219203733.3063192-1-elliott@hpe.com> References: <20221219203733.3063192-1-elliott@hpe.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Proofpoint-GUID: 8tTJD82xn1hsm32dpy-VUM8DmsXSWayU X-Proofpoint-ORIG-GUID: 8tTJD82xn1hsm32dpy-VUM8DmsXSWayU X-HPE-SCL: -1 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.205,Aquarius:18.0.923,Hydra:6.0.545,FMLib:17.11.122.1 definitions=2022-12-19_01,2022-12-15_02,2022-06-22_01 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 phishscore=0 spamscore=0 clxscore=1015 suspectscore=0 impostorscore=0 malwarescore=0 lowpriorityscore=0 adultscore=0 priorityscore=1501 mlxlogscore=999 mlxscore=0 bulkscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2212070000 definitions=main-2212190181 Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Always yield to the scheduler at the end of skcipher_walk_done(), not just if additional bytes are left to be processed. This avoids soft lockups if drivers are invoked back-to-back to process data that is an integer multiple of their block size. Example: while processing 1 MiB buffers, multiple skciphers run from 192 s to 218 s without ever yielding to the scheduler, causing three soft lockup complaints. The kernel is configured for CONFIG_PREEMPT_NONE=3Dy (or preempt=3Dnone on the kernel command line), so only explicit cond_resched() calls trigger scheduling - might_resched() and preempt_enable() do not (see kernel/sched/core.c). [ 190.865601] tcrypt: PERL my %speeds_skcipher =3D ( [ 192.041157] tcrypt: PERL "cbc-aes-aesni" =3D> 2396490, [ 192.373934] tcrypt: PERL "ctr-aes-aesni" =3D> 574888, [ 193.548967] tcrypt: PERL "cts-cbc-aes-aesni" =3D> 2396018, [ 193.909077] tcrypt: PERL "ecb-aes-aesni" =3D> 631824, [ 194.223801] tcrypt: PERL "xctr-aes-aesni" =3D> 535778, [ 194.608548] tcrypt: PERL "xts-aes-aesni" =3D> 676518, [ 196.440247] tcrypt: PERL "ctr-aria-avx" =3D> 3804224, [ 196.675788] tcrypt: PERL "xchacha12-simd" =3D> 368668, [ 196.988868] tcrypt: PERL "xchacha20-simd" =3D> 535314, [ 197.301510] tcrypt: PERL "chacha20-simd" =3D> 535142, [ 198.590113] tcrypt: PERL "ctr-sm4-aesni-avx2" =3D> 2642930, [ 208.975253] tcrypt: PERL "cfb-sm4-aesni-avx2" =3D> 22499840, [ 218.187217] watchdog: BUG: soft lockup - CPU#1 stuck for 26s! [modprobe= :3433] [ 246.181238] Modules linked in: tcrypt(+) hctr2 essiv adiantum ... [ 246.185048] RIP: 0010:measure_skcipher_bigbuf.constprop.0.isra.0+0x282/0= x393 [tcrypt] [ 246.185304] Code: 00 0f 31 ... ... [ 218.197313] Call Trace: [ 218.197567] [ 218.197822] ? 0xffffffffc052a000 [ 218.198079] do_test.cold+0x989/0xc7a [tcrypt] [ 218.198340] ? 0xffffffffc052a000 [ 218.198595] tcrypt_mod_init+0x50/0x1000 [tcrypt] [ 218.198857] ? 0xffffffffc052a000 [ 218.199112] do_one_initcall+0x41/0x200 ... [ 219.391776] tcrypt: PERL "cbc-sm4-aesni-avx2" =3D> 22528138, [ 221.560847] tcrypt: PERL "ctr-sm4-aesni-avx" =3D> 4560732, [ 231.960414] tcrypt: PERL "cfb-sm4-aesni-avx" =3D> 22498380, [ 242.350070] tcrypt: PERL "cbc-sm4-aesni-avx" =3D> 22527668, [ 244.471181] tcrypt: PERL "ecb-sm4-aesni-avx" =3D> 4469626, ... [ 246.181064] watchdog: BUG: soft lockup - CPU#1 stuck for 52s! [modprobe= :3433] ... [ 250.168239] tcrypt: PERL "cbc-camellia-aesni-avx2" =3D> 12202738, [ 255.840094] tcrypt: PERL "cbc-camellia-aesni" =3D> 12203096, [ 264.047440] tcrypt: PERL "cbc-cast5-avx" =3D> 17744280, [ 273.091258] tcrypt: PERL "cbc-cast6-avx" =3D> 19375400, [ 274.183249] watchdog: BUG: soft lockup - CPU#1 stuck for 78s! [modprobe= :3433] ... [ 283.066260] tcrypt: PERL "cbc-serpent-avx2" =3D> 21454930, [ 292.983848] tcrypt: PERL "cbc-serpent-avx" =3D> 21452996, ... By adding a unilateral call to crypto_yield(), which calls cond_resched() and lets the scheduler use the CPU for another thread, that no longer happens. Starting at 2218 s, there is no soft lockup reported at 2244 s: [ 2217.202692] tcrypt: PERL my %speeds_skcipher =3D ( [ 2218.450215] tcrypt: PERL "cbc-aes-aesni" =3D> 2179138, [ 2218.950960] tcrypt: PERL "cbc-aes-aesni" =3D> 538738, [ 2219.460618] tcrypt: PERL "ctr-aes-aesni" =3D> 575212, [ 2219.983006] tcrypt: PERL "ctr-aes-aesni" =3D> 574402, [ 2221.329550] tcrypt: PERL "cts-cbc-aes-aesni" =3D> 2182864, [ 2221.840599] tcrypt: PERL "cts-cbc-aes-aesni" =3D> 539064, [ 2222.344290] tcrypt: PERL "ecb-aes-aesni" =3D> 537402, [ 2222.869201] tcrypt: PERL "ecb-aes-aesni" =3D> 537730, [ 2223.400315] tcrypt: PERL "xctr-aes-aesni" =3D> 534824, [ 2223.897915] tcrypt: PERL "xctr-aes-aesni" =3D> 534782, [ 2224.414956] tcrypt: PERL "xts-aes-aesni" =3D> 539592, [ 2224.923715] tcrypt: PERL "xts-aes-aesni" =3D> 539356, [ 2226.740211] tcrypt: PERL "ctr-aria-avx" =3D> 3392444, [ 2228.545624] tcrypt: PERL "ctr-aria-avx" =3D> 3392068, [ 2228.869883] tcrypt: PERL "xchacha12-simd" =3D> 368932, [ 2229.204980] tcrypt: PERL "xchacha12-simd" =3D> 374122, [ 2229.609975] tcrypt: PERL "xchacha20-simd" =3D> 535596, [ 2230.022425] tcrypt: PERL "xchacha20-simd" =3D> 537500, [ 2230.429674] tcrypt: PERL "chacha20-simd" =3D> 535474, [ 2230.831041] tcrypt: PERL "chacha20-simd" =3D> 534264, [ 2232.278150] tcrypt: PERL "ctr-sm4-aesni-avx2" =3D> 2640770, [ 2233.744781] tcrypt: PERL "ctr-sm4-aesni-avx2" =3D> 2642520, [ 2244.290542] tcrypt: PERL "cfb-sm4-aesni-avx2" =3D> 22497308, [ 2245.725044] tcrypt: PERL "cfb-sm4-aesni-avx2" =3D> 2604468, [ 2256.279228] tcrypt: PERL "cbc-sm4-aesni-avx2" =3D> 22526084, [ 2257.729868] tcrypt: PERL "cbc-sm4-aesni-avx2" =3D> 2600460, [ 2260.068782] tcrypt: PERL "ctr-sm4-aesni-avx" =3D> 4560650, [ 2262.414663] tcrypt: PERL "ctr-sm4-aesni-avx" =3D> 4561468, [ 2272.943000] tcrypt: PERL "cfb-sm4-aesni-avx" =3D> 22496026, [ 2275.233755] tcrypt: PERL "cfb-sm4-aesni-avx" =3D> 4456984, [ 2285.779516] tcrypt: PERL "cbc-sm4-aesni-avx" =3D> 22525908, [ 2288.081160] tcrypt: PERL "cbc-sm4-aesni-avx" =3D> 4457036, [ 2290.374086] tcrypt: PERL "ecb-sm4-aesni-avx" =3D> 4465790, [ 2292.677381] tcrypt: PERL "ecb-sm4-aesni-avx" =3D> 4466014, [ 2298.544718] tcrypt: PERL "cbc-camellia-aesni-avx2" =3D> 12246268, [ 2299.869611] tcrypt: PERL "cbc-camellia-aesni-avx2" =3D> 2349440, [ 2305.734078] tcrypt: PERL "cbc-camellia-aesni" =3D> 12246930, [ 2307.746065] tcrypt: PERL "cbc-camellia-aesni" =3D> 3832992, [ 2316.127414] tcrypt: PERL "cbc-cast5-avx" =3D> 17737348, [ 2318.703437] tcrypt: PERL "cbc-cast5-avx" =3D> 5061014, [ 2327.694881] tcrypt: PERL "cbc-cast6-avx" =3D> 19065488, [ 2331.672188] tcrypt: PERL "cbc-cast6-avx" =3D> 8145590, [ 2341.750274] tcrypt: PERL "cbc-serpent-avx2" =3D> 21453172, [ 2343.209420] tcrypt: PERL "cbc-serpent-avx2" =3D> 2611702, Fixes: b286d8b1a690 ("crypto: skcipher - Add skcipher walk interface") Signed-off-by: Robert Elliott --- crypto/skcipher.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/crypto/skcipher.c b/crypto/skcipher.c index 0ecab31cfe79..cdead632117a 100644 --- a/crypto/skcipher.c +++ b/crypto/skcipher.c @@ -153,13 +153,20 @@ int skcipher_walk_done(struct skcipher_walk *walk, in= t err) scatterwalk_done(&walk->in, 0, nbytes); scatterwalk_done(&walk->out, 1, nbytes); =20 - if (nbytes) { - crypto_yield(walk->flags & SKCIPHER_WALK_SLEEP ? - CRYPTO_TFM_REQ_MAY_SLEEP : 0); + /* + * Allow scheduler to use the CPU since it has been busy, + * regardless of whether another loop pass is due + */ + crypto_yield(walk->flags & SKCIPHER_WALK_SLEEP ? + CRYPTO_TFM_REQ_MAY_SLEEP : 0); + + if (nbytes) return skcipher_walk_next(walk); - } =20 finish: + crypto_yield(walk->flags & SKCIPHER_WALK_SLEEP ? + CRYPTO_TFM_REQ_MAY_SLEEP : 0); + /* Short-circuit for the common/fast path. */ if (!((unsigned long)walk->buffer | (unsigned long)walk->page)) goto out; --=20 2.38.1