From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mout.web.de (mout.web.de [212.227.17.11]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 078E03BB13D; Mon, 10 Aug 2026 11:27:52 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=212.227.17.11 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786361275; cv=none; b=eJY5mNi2XcB/1+uUSnpKUdX2UC1sQe0fhza+DuV3jhJwdNu+2A1PDc8xmQrjzCXLBkItch0hEA/+SVhkKi2aMuUXA5KwvM+aFWft3GCqPkuWXkS+XYvd0FcrVwOXG70OmqpsXR72wcIKYnUVLyRPyLwuUPx78uws6eXAfY3hyRc= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1786361275; c=relaxed/simple; bh=Fbw2ViJBbHmwuosEwB4v+imM+9hYQN3AqEVJ9cKaawQ=; h=Message-ID:Subject:From:To:Cc:Date:In-Reply-To:References: Content-Type:MIME-Version; b=RrTKH0pZsUsF7nXLXV8F96klljKmLCXXzT0HEgGnOHkSwiUxWD7HrojcR1nPsj4XJLI/Jcz2MDNPdAovRCBOEW6LzIurAwoeIiKR37SLWq9vAYiIOS/ZIfIdPQiqe+cchPPY83NiYGxmctcYb/9FNW9JRBisP5ys95ez+pQtvz8= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=web.de; spf=pass smtp.mailfrom=web.de; dkim=pass (2048-bit key) header.d=web.de header.i=spasswolf@web.de header.b=FB5rR/+5; arc=none smtp.client-ip=212.227.17.11 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=web.de Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=web.de Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=web.de header.i=spasswolf@web.de header.b="FB5rR/+5" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=web.de; s=s29768273; t=1786361270; x=1786966070; i=spasswolf@web.de; bh=b46wsTJjy/Jjj97yBvBbzCH/gQQvVj+KrXCTFD/HQHU=; h=X-UI-Sender-Class:Message-ID:Subject:From:To:Cc:Date:In-Reply-To: References:Content-Type:MIME-Version:Content-Transfer-Encoding:cc: content-transfer-encoding:content-type:date:from:message-id: mime-version:reply-to:subject:to; b=FB5rR/+5Hu37h6JNHRK+ctB84vUpCuUZSuLfxbewww5mZj8pq4FQDlAnvayzYuxv pAbDdi6iYOM73iYdGeo88kaeD/Iy0pBMGw4PjCgFZ78yZVqt7ovnNj+u2i/jzDvBh IOejRwBbdo8Py0RGTFN+5cNGAFbWr0smLI3Ggm9UiHJPWaUlu1XfW4OaAm3iE390d eO85TaRlY8iMJ3jo+pVWNAPXIUD94XuoX+GePn2fPOoU9jBGkuw5pcqEx+QvwU9ZW lMbW3nwDhj1qx/7YdQg+eXhH6gClaj9Tkon0T3AaG/usZsno5twIfB566dRaM4xYA InWhFd5B3vWTgWFJFA== X-UI-Sender-Class: 814a7b36-bfc1-4dae-8640-3722d8ec6cd6 Received: from client.hidden.invalid by smtp.web.de (mrweb105 [213.165.67.124]) with ESMTPSA (Nemesis) id 1MQPdr-1wXALn2ulV-00LXk2; Mon, 10 Aug 2026 13:27:50 +0200 Message-ID: <6e837ba2419216d9d142766a13f3947fbd456dfd.camel@web.de> Subject: Re: [Re] GPU reset when running the ROCm hsa runtime tests on gfx12 and next-20260701 From: Bert Karwatzki To: Alex Deucher Cc: linux-kernel@vger.kernel.org, amd-gfx@lists.freedesktop.org, linux-next@vger.kernel.org, Jesse Zhang , Amber Lin , Mario Limonciello , spasswolf@web.de Date: Mon, 10 Aug 2026 13:27:49 +0200 In-Reply-To: References: <20260705003504.31425-1-spasswolf@web.de> Content-Type: text/plain; charset="UTF-8" User-Agent: Evolution 3.56.2-10+b1 Precedence: bulk X-Mailing-List: linux-next@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Provags-ID: V03:K1:JOXdFEdxQ20qpDtYFBmK1S5bLpZ70NPOw4lw6lsHh3jx6Hm7Wrq IQ8Pc5hcG+XtYgE3XPfJnTXf84BX1RvO+Zds9ww3pwu8RvkMBds3NfuKAKKduGFomanWTKW dMX2JaVU7aDGjDo2j4/f/Dtg1lVMqScTXeABH+UBi697+vjwRDrxjzI5za2Q8UahH/D8GJ1 kPyXrpbCWsqAJ4AtE7Qlg== X-Spam-Flag: NO UI-OutboundReport: notjunk:1;M01:P0:7wqwqqM7WR4=;Lgg2Z7I7b5xK2W6HUnIKlMP4JiB cu6EhKqr8PquZyzV04MTgUI35m/rV/wc0fgszZs2bHkwIN/ky4VCAW5ibJ/famuLYsxD5v/S2 bJT8fOsDPP9LYA4+I8+3OZxpSlqtw58dYrZD0hPT8GGETnr9i9jUL06FlsUSEjcgBR1K6KAne X4OGrDQmmjHYCMeRA2ajyXS6Youq1C3d9zuI3eICMeD2ofFV4eehpXSWB/Av5FDUrYi0u5+o4 KYrVapnzlLRTiRc3B9RWwc7js5Jzc3Z2id30lBQ7d80d7DVnmm3k//gGf/QiDc56wZ3o4diRa YbZv5Me2Z99qXbm/kgoXCTy7E8oPa+5dWyKKkxTcPizvi3j9a9/9ab3e7fJfq1u9PfMlagzwC xUHZPqEm1pyg1IHdj4bojvJSSGx6r1zhRPhf+vSDtSRHd8uLKE2/cGbOcO7dLSW86tKq2yt5R ovKbalYF5jg3fTDnw8mCDNsdI5MchsFv3UJ9HTzJkKPBIRjL/vvXtdtcu9A9B1YiwpS3t0Vp/ hSEf90i74gRR0J0E/5Qf1vcBRAhOApiQBRZq3SzhYmCI9VgZvzoU0S2b1rHLky6LtBvVKrJjj DR7YLcSM/PocL8pbBEVUEMZi5cUZ/+B0N8vi7GudlR6L3ElYBOxq1CoCg+jumA67B9ymG1Ryz 3B+9YHl2Ct5gqEh2vuzcSNtAkr4qxdFAwZcZmUjzvvOPsTeO4MIvWoQ5GIHXqMx1G3FWVelgP 5ktMigaJfWaMqNVB1yaIMRFrJcwEv4Z1AmWCXCELRG/XyoGbA+ElZ1bYZKUe0h6p2SMBliVxZ vtecOvn6tcKMrjsmgpTC/FBpV4qZrQWdvPKs1MUfEKprkdA+kyjOM9vvUUw/e2/0PSPHY3Kd0 pGJgDaLcQbaXAVzURprcf0i5zqiPX6nUQtIaKJ6DI6fkCYY8SJLsPeQ+8N4UPVrWFDl7veWjV mz+L418ss+5ohm5jxY+bZgXxsKslZeISOO5F3RCL/pP02fJndtaHIHvN8+sWYOwhK7BVIiG0X PtAOSQZWbioJKopwjRsYPIyq+32omlHMnaSvV1CTiXHpMxeXqYGfjdTYY0E0ta21mzc2rzrpa auyx9X3/CoUnMloOYFAvlJ3h8W2k9r1Ah0bv/dznV0pOeEh2tA7wH2ll1tk4M1RBguWybBMvp 3h8wcW3h0a5oF2jUwZHVlQzePf6eImVBw0IirVy3EF2UXJAg/DJKsr/3DxIT068djy0AZAaij E37XCkWfPjWpQlNjIP0esOSXcBKOy4zpQzrui4hMesJC6cDBQSQ69NDWgfi2Ye7GPMjXGTl7b K68hm8N+2JGrkQ4FsZHVjfNOHhNaT1SdUB6ocElT2wwrPqz3f2Z+DkPlR0zfESle7LwoDBLrQ Ex+QtFsTolf5G5wKBdmHpjdmKh4iVTGjFNQUy0BehvHjlEP9dy37GlTsfNBUNo2UTdbI7mroF zR27DzDzxXWQNKdxzSVw22cfh1Pi45k2TrRnGkXUcxrzIpVxVqN7oosHhW4zoK9TXX4+TRd3t YNDWUC3orkEJ9Ny/K+t9iWlGpPd4aD/0bkNRmz9l8ecwnMS2wj4uz9/VigL3N6hUJYxXD7xgi vV+3SQvksdJTruyDNnjd/C7ywHB6w045OO+ecLgEVI2UWTKC4lG53x/PLusSnsDhD9clzdt3x W2xLRvpGq+b0zagKa8dpEKXgyeuNwIokiXAkoCWVuthigckrQ1FxFsZAckGE1uu0uF6684WM4 NUXpc66SHkzD9Xv8N93u9Ozaj888wjTr5x3aOfJqtaulz7AQBN3UWJiCVhRzviJN6ha0rMo6p WRWhwDnccww80wMzrhnP6HUceKF8NdEocLKHI3BblcpP6uBQ3eOFrX3LbkCOXy/6F3ki+l8aF 2Yc2Gg4DVUnNqmNxav4NjRtmHZJia1LRcuv0hyrnnNDwJFF4sSmufKhTcHpSsjkGjarYue6OZ P85TtgYH9vyCXYeI5Szl0uqYlEfVgAFAnUU7R3NyMvEK8vTYnFlPqV9booKQK49ka05F38hjZ GNVr8fOOLa/Xlz6iNI5ZvPza801tLyxEFb7vp8hzcd7XQvLN7obkQnnHC4ZB0TlDsqx70a5Dt oLdQA3esToUfDrmxNqmQo9LrfFQIxUXeV3ZNvcpR/DerlZCCsokTfVdBKC9HtC7s2xJYfTaLg 2VK9KQ9kGU6g8z0Fk8QrrH0NtI8uMcpEFoeCk0vDmyX7Qo3w6exmWT6y9Hus+z2ZW+F8lu62t icNXBr5jCJ3CMEqebisd9bE2vxXg09qsEsYQCf0JNJtf+5iz+X276Kjy+BxZ6MSVO4AZ3MbFp uclRdDPojG+xLHk9BfgILQ3UFieFtHxswrmrdiNNZBykbeYWB52TSXOVPd5Gze/1PgUBZvAUc lWV+GO8A6vKS80/e8kv3BXQe42aDG+tiyA9lVnQ9elyntnOWOY5qVyLg8ja/MaiYjAQ1tF8xN wfhVwBrcLmkbt0BFUZmnFC1ixEfVhuG0FwNgWEyY5fvIdsJMyvmG00UGa2rkI1WWl9BKCMm45 5Gx+pDVv9w27lWApks2LzKVVAhSUQU0pxuv/so4rRdEVaScNDvh4xftseq4vrwtM0MflHQD+p WCfLDPUTrdDeEy8G3DLHgtx/rBDZFXYO05c3uBo4uYCfcDL74TnmU0ybFjYwWQPAZEZt7mdtp AZx9UdoZtGrJIEB+sZoEAX0iBf5NkKlOoti0BjWX4XjvHW2PRaa1lUrgzsM/HI8uyqzmeIBBR hHQ2vbpCURfooMrQwV06qN06TVRLya7/WE216ZpQfJ08fzaTZOMAjztMv3Ej2Ck85M+sob69n +SvIRJPK7I3hQQRn8oENU/3X8cKnONLiq8YjNhZ9NAzb9Ik55Ztq48AGeu3OsZ1KCN3Nwmt/o zaXh/OFbZNKPd8/rs+qFH9x2Nin3vYq++jMXCYNROmb8FV+OgaHwdWXy1SUBdIXWVcGK7tXks PwmBCs/pDno2ZKVsWdKfjd92Flwj4bTlrn4FCaJTwHHQg6P3+vjiYoLPxOhQszsovxEqLB8n3 RJENUodXPN5HTFR03vcBpMLiJrSwDo0mVAhikKD8qiEnzVr2eMXr8uAYoVLeKRgjwaUghenPo fVvsRidQMzIWwu/893fUHiLGRhJx+M5rZr7fhwxwcJ816z64It5accN+OMPeLUiVSGf21xUBa rucFJQfqLszJ68IzZiE9R1I1GMlDyKpnsBsGiq5Q3WNLzIvKRRvP6mLE5Qhni2OHyqYxb3nOS 5giP1oG16hmMrd6g/Oquxzx6VIofzHOhWnPZ4DfKYMV/W66rW05OhD/uVLazcNUPogjD7EXw7 3PNStGhWUNONG6/Pc1lYgMt4Npa2OjcZzlhd7alQpngLNGzBMWJ2kNq2dNRriraw9ZMU9DbLo Yf55q46YuETTCqa75cUwIL7qhhi2X8KoEd4dbG8GcraELg9cYujQnt/51wAlgsEikEIxoXZD3 NEEUQVGqL8W8TmV6wKGrX9BO5ivdp2VcsPREydSOXFtg+r/V5hNNcxUlxTw5ad24Fg9YUL+V1 JjXL+g1rxJnytHh1kMjpcP3nUk0h+EcAEnW9LZSYnaLC9xSOKkcQJQPAAZBbAhrBIEpHJrJXT qeT1atreP1FegSrqsBOKNyyeI8pWssPqEkURyeU6SSx12eaW9p0tATcBD24ca6pAr0GFQtefv /OM9zb5WMbmN83NQqnMOxdJeYD07T8GiACxiYolVF71NsGsLl/i8VRYQpThOjupnLomCv0epb p1he3SnpavpQwUwH9U/h8M7IWojST/QCZMsBRKIc4aGlIM8L5zOx4DR0PJF23MZDBi5J0tgpx HZ6R8WW12FdIwSCEs+Anb90PxUBB9yJm0Tx6ve3Og51vtCTFQ2ZjbOvYXCxrbD1Q3uNVPgLSW +cjcwNYsup04Fgy5/ykT9NtzETF2lDk/Vi3sYJMpFlKg5ENos5JMwfJesoFCDtaFrqqefrGXI y+MNAnyBIG1pORfiQBaEuLlUK7LiYvR80EuneR1toH4TdQIEA+81K31iZYQPzLapQrgyGWDsU P7FW50zaQhhmA3TKpiFGbweBsh/gmWE86x6XuY/o4NSq4TaweoSFF6yjzK07sJtC9/gn8zE/P catdDfH7/pHFDcseYpLZ8/CqQOxelAXN92DYdrCYJA6YqFx3YfwYwwuqLGVsIxm82C/vLQK1p c0VF8mqwnTlrvEzdtTM/3pRfm3wTpj36tdH8gyGPQk6Ys2FYfjBpVyCRayRy17yGep0WhQ/Gu eDSclXhX4bkC64GjJ51XqJ4ami7V3KQ3YzQHqcLD+vFEAXEaLeHqSjoW+iRVa4LEEdn5xzFVA j4Kjel0f1mCszGplnftfDCd7Af1wKf/iresFK/IR1MonbEsJPJrtg9ykzrJufZIFmhvC/4G5I /Vki0/1PDsO/DNCa5NAY7m1D+xNQgPduCEqEbiFLvhCGhnkwRHaGLzSNrL+zhN558+1KeXmyn FdUKpVYI452ui2U9pkKk13DJZtw7ZQ394eHlLZwXiUvtzV/HvT+cRpRoWYU4nscd4asJTIuZ+ tEIixOJtu5pW2w5mVdrU4/+g8oc1dMcW9BPtAJCJTgguvX1xzOCJzTTKzuf5tJkszW62Jv35G j6D4EB3hXP9GC7aMECGAPJTDrd6VQ733+pyK7cH/1/MHP0ktj82qj/4ihWU66t+UlyDmRnI2W Ee/lN3vlpIZEkbFgnISb76u3vlL6uK6uroTmU6rfuJXGjWGLXFnMIcxtvE4Xsf4gvZYcLCUTY o3jIHs3ObX9PoYCDZGkkb0d4tg9+UGuiFZsxUq/fizY3gSCxVBsLRJ8Jc7v4IvTcxsD8JQrCT XdBcnpNzYoZzcNGxMXJi0HnH8KrOad1KuyF3UuNu/ZKE+FIH9k5hy2XLTAPJAALsdRLiME12g pl3HBO5cjHVrhZYRZIupC/vDCeF4dhMtj8gJfR59uTLwJxZ2u8qvLRZifMIg/AmI09K9P66ep 738UseSvtGB+UN+GCwEKrg5q4t7f9ffNnPNTim61Lc80iSUM57KjZucO2wScmoBIpM6XYvnU5 9VbiSzplhneyMGqSmTT7sGB8IpYewo3FraUPM83V1lNlPizK+ZstSyakiNeoZDLUgAZ752Q1g Vclj2gsCh1DizR55Q2SUjIRoYuR+KuB4bMFBElmKo3eHnP4PkK89g9K7XEpU+2sJJBRe/4PDQ XTg+C3VRnsNQL0hPfRGmbt0LxDonBzjBjWM2VG3zaBR7y3z02qlyucC8z9NzjnQYtdSEcdI+r 1eKc+efvZh9y2bpcJTwrpNgzgrzS/R3aZCLm0hUGrL0HXhNyuW8qEszIIKLK65umcy17tMnG5 XeCjkWSGVUCSACiXy7qG9LsUAzx/tGGWBjD83iortScrvrxUw4TnJomkTzVHzvW+M9weynxhH B9ZlO8ihpZlxeEWsLKyxjHHwxmtgZEL5tb78EeRPsGsDKB7oF7ic8ZwByDNrXXVnY4L2WT+1M +jva75yTWLx6IG1wRdabwyf8i0rF+NokG3oRzlx8GcXZ6tu1aox0X9C/vGMJJo4cPsDWPR9FJ kknKA9dogQvlmL/yWWhU7eOwf5CBaleUR5cRdbuyJM1m7cP79xBa/9YznHNabxp35LvBLX8Xs cX3lIzLKxCTUXeSqePzmJFxk2MQRS3483P7laBvHXW8j+aAjOur9GbqpQI7f1TcJ8oGsS0W7p DvRSWL1BCCu2M9uRjXC09fW+n2BwCti8kg7YPoaZvYlcrhv5TZ2BwOWXkC+FFSdSBl9HoHILl eKIpO47VGTAmOdhE436g3df56k3BdqoFFDLnD5SVJU9/OyvtWMZYMUQwbeoscvcIfXJHhRmmW H0VhvyYRZKbe2cvU8phpMMrbn6kB2Bg9NaH/uwdYAaaI7jR5ERixNgJrID//WUJUXpKFivsLQ AUBfH9+eQnRSMXRa4q5GGx8fUbcc3xuUyHVjnolTGrLA6MbnOxkE2PZG5M/epdS6iEI/lDiDB soEfbpLV5VlDW8MyjZTCVG4+w1JNrA3jtMB1+H70bAXAGeStGvKXWwzwO2PC/ZrfR88AmJwYD 1MUtEh/IIPepOzWeS3KTsfdVvuJAHNsBmzOo3cFBhLpsb85qOnyq4tAJLuCanbXekQ78etmGe 5xmAzD9OwlAnO+6Wrn7SYkY9kJPyoFXlg/vaayxhi3UmhM2y8F6DSU+y6UL85Q== Am Montag, dem 06.07.2026 um 14:46 +0200 schrieb Bert Karwatzki: > I found the real cause of my problems: >=20 >=20 > bool amdgpu_mes_queue_reset_by_mes_supported(struct amdgpu_device *adev) > { > u32 ip_maj =3D IP_VERSION_MAJ(amdgpu_ip_version(adev, GC_HWIP, 0)); > u32 ip_min =3D IP_VERSION_MIN(amdgpu_ip_version(adev, GC_HWIP, 0)); > u32 mes_sched =3D adev->mes.sched_version & AMDGPU_MES_VERSION_MASK; > printk(KERN_INFO "%s: ip_maj =3D %u ip_min =3D %u mes_sched =3D 0x%x", = __func__, ip_maj, ip_min, mes_sched); >=20 > return (ip_maj =3D=3D 11 && mes_sched >=3D 0x8c) || > ((ip_maj =3D=3D 12 && ip_min =3D=3D 0) && mes_sched >=3D 0x8d) || > ((ip_maj =3D=3D 12 && ip_min =3D=3D 1) && mes_sched >=3D 0x73); > } >=20 > returns false on my machine (because mes_sched is not large enough) >=20 > [ T8549] amdgpu_mes_queue_reset_by_mes_supported: ip_maj =3D 12 ip_min = =3D 0 mes_sched =3D 0x76 >=20 > So I skipped the call to amdgpu_mes_queue_reset_by_mes_supported() >=20 > diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/dri= vers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c > index 5c9dfb0c424f..462f20aeb681 100644 > --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c > +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c > @@ -448,17 +448,22 @@ static int reset_queues_mes(struct device_queue_ma= nager *dqm, struct queue *q) > unsigned int num_hung =3D 0; > int r =3D 0; > struct mes_remove_queue_input queue_input; > + printk(KERN_INFO "entering %s", __func__); > =20 > + /* > if (!amdgpu_mes_queue_reset_by_mes_supported(adev)) { > r =3D -ENOTRECOVERABLE; > + printk(KERN_INFO "%s: reset by mes not supported", __fun= c__); > goto fail; > - } > + }*/ > + printk(KERN_INFO "%s: skip calling amdgpu_mes_queue_reset_by_mes= _supported()", __func__); > =20 > /* reset should be used only in dqm locked queue reset */ > if (WARN_ON(dqm->detect_hang_count > 0)) > return 0; > =20 > if (!amdgpu_gpu_recovery) { > + printk(KERN_INFO "%s: gpu recovery not enabled", __func_= _); > r =3D -ENOTRECOVERABLE; > goto fail; > } > @@ -470,6 +475,7 @@ static int reset_queues_mes(struct device_queue_mana= ger *dqm, struct queue *q) > queue_input.xcc_id =3D ffs(dqm->dev->xcc_mask) - 1; > /* pass the known bad queue info to the reset function */ > r =3D amdgpu_gfx_reset_mes_compute(adev, NULL, NULL, NULL, &num_= hung, &queue_input); > + printk(KERN_INFO "%s: amdgpu_gfx_reset_mes_compute() returned %d= ", __func__, r); > if (r) > goto fail; > =20 > @@ -3231,6 +3237,7 @@ int kfd_dqm_suspend_bad_queue_mes(struct kfd_node = *knode, u32 pasid, u32 doorbel > struct qcm_process_device *qpd; > struct queue *q =3D NULL; > int ret =3D 0; > + printk(KERN_INFO "entering %s", __func__); > =20 > if (!pdd) > return -EINVAL; > @@ -3242,6 +3249,7 @@ int kfd_dqm_suspend_bad_queue_mes(struct kfd_node = *knode, u32 pasid, u32 doorbel > =20 > list_for_each_entry(q, &qpd->queues_list, list) { > if (q->doorbell_id =3D=3D doorbell_id && q->prop= erties.is_active) { > + printk(KERN_INFO "%s: calling recover_ba= d_queue_mes() for queue %px", __func__, q); >=20 > and got this output when running the hsaruntime (which inject illegal op= codes into the > command stream) test: >=20 > [ 113.811612] [ T645] [drm:gfx_v12_0_bad_op_irq [amdgpu]] *ERROR* Il= legal opcode in command stream > [ 113.811675] [ T2558] entering kfd_dqm_suspend_bad_queue_mes > [ 113.811676] [ T2558] kfd_dqm_suspend_bad_queue_mes: calling recover= _bad_queue_mes() for queue ffffa03160950400 > [ 113.811676] [ T2558] entering reset_queues_mes > [ 113.811677] [ T2558] reset_queues_mes: skip calling amdgpu_mes_queu= e_reset_by_mes_supported() > [ 113.811887] [ T2558] reset_queues_mes: amdgpu_gfx_reset_mes_compute= () returned 0 >=20 > So even though amdgpu_mes_queue_reset_by_mes_supported() reported false= , > amdgpu_gfx_reset_mes_compute() returns 0, suggesting that resetting actu= ally works here! >=20 > Perhaps the minimum required mes_sched version for 12.0 can be relaxed t= o 0x76 to solve this: >=20 > From 6d7af652177063963012eb4df228e99caeb03b31 Mon Sep 17 00:00:00 2001 > From: Bert Karwatzki > Date: Mon, 6 Jul 2026 14:36:22 +0200 > Subject: [PATCH] amdgpu: relax required mes_sched version >=20 > This mes_sched version is actually enough on this hardware: > 03:00.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [= AMD/ATI] Navi 44 [Radeon RX 9060 XT] [1002:7590] (rev c0) >=20 > [ 3.200538] [ T537] amdgpu 0000:03:00.0: initializing kernel modes= etting (IP DISCOVERY 0x1002:0x7590 0x148C:0x2437 0xC0). > [ 3.200546] [ T537] amdgpu 0000:03:00.0: register mmio base: 0xDFC= 00000 > [ 3.200547] [ T537] amdgpu 0000:03:00.0: register mmio size: 52428= 8 > [ 3.203739] [ T537] amdgpu 0000:03:00.0: detected ip block number = 0 (soc24_common) > [ 3.203740] [ T537] amdgpu 0000:03:00.0: detected ip block number = 1 (gmc_v12_0) > [ 3.203741] [ T537] amdgpu 0000:03:00.0: detected ip block number = 2 (ih_v7_0) > [ 3.203741] [ T537] amdgpu 0000:03:00.0: detected ip block number = 3 (psp) > [ 3.203742] [ T537] amdgpu 0000:03:00.0: detected ip block number = 4 (smu) > [ 3.203742] [ T537] amdgpu 0000:03:00.0: detected ip block number = 5 (dm) > [ 3.203743] [ T537] amdgpu 0000:03:00.0: detected ip block number = 6 (gfx_v12_0) > [ 3.203743] [ T537] amdgpu 0000:03:00.0: detected ip block number = 7 (sdma_v7_0) > [ 3.203744] [ T537] amdgpu 0000:03:00.0: detected ip block number = 8 (vcn_v5_0_0) > [ 3.203744] [ T537] amdgpu 0000:03:00.0: detected ip block number = 9 (jpeg_v5_0_0) > [ 3.203745] [ T537] amdgpu 0000:03:00.0: detected ip block number = 10 (mes_v12_0) >=20 > Signed-off-by: Bert Karwatzki > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) >=20 > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c b/drivers/gpu/drm/a= md/amdgpu/amdgpu_mes.c > index 6c0dde3786e3..c88fdc8a187d 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c > @@ -869,7 +869,7 @@ bool amdgpu_mes_queue_reset_by_mes_supported(struct = amdgpu_device *adev) > u32 mes_sched =3D adev->mes.sched_version & AMDGPU_MES_VERSION_MASK; > =20 > return (ip_maj =3D=3D 11 && mes_sched >=3D 0x8c) || > - ((ip_maj =3D=3D 12 && ip_min =3D=3D 0) && mes_sched >=3D 0x8d) || > + ((ip_maj =3D=3D 12 && ip_min =3D=3D 0) && mes_sched >=3D 0x76) || > ((ip_maj =3D=3D 12 && ip_min =3D=3D 1) && mes_sched >=3D 0x73); > } > =20 The specific hsaruntime tests which cause gpu resets on a RX9600 (gfx1201)= on next-2020701+ are $ /usr/libexec/rocm/libhsa-runtime64-tests/run-tests --gtest_list_tests | = grep Invalid Queue_Validation_InvalidDimension Queue_Validation_InvalidGroupMemory Queue_Validation_InvalidKernelObject Queue_Validation_InvalidPacket Bert Karwatzki