| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <202203120337.51TmNkmV-lkp@intel.com>
Date: Sat, 12 Mar 2022 03:02:08 +0800
From: kernel test robot <lkp@...el.com>
To: Andrey Grodzovsky <andrey.grodzovsky@....com>
Cc: llvm@...ts.linux.dev, kbuild-all@...ts.01.org,
linux-kernel@...r.kernel.org,
Christian König <christian.koenig@....com>
Subject: [agd5f:amd-staging-drm-next 1275/1333]
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:5089: warning: expecting
prototype for amdgpu_device_gpu_recover(). Prototype was for
amdgpu_device_gpu_recover_imp() instead
tree: https://gitlab.freedesktop.org/agd5f/linux.git amd-staging-drm-next
head: d974d4a696618e82ed6d39ad2b3688d915bb13dd
commit: 92fc842d9da15ae9b14dc964d115e250c7b568e1 [1275/1333] drm/amdgpu: Serialize non TDR gpu recovery with TDRs
config: riscv-randconfig-c006-20220310 (https://download.01.org/0day-ci/archive/20220312/202203120337.51TmNkmV-lkp@intel.com/config)
compiler: clang version 15.0.0 (https://github.com/llvm/llvm-project 276ca87382b8f16a65bddac700202924228982f6)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# install riscv cross compiling tool for clang build
# apt-get install binutils-riscv64-linux-gnu
git remote add agd5f https://gitlab.freedesktop.org/agd5f/linux.git
git fetch --no-tags agd5f amd-staging-drm-next
git checkout 92fc842d9da15ae9b14dc964d115e250c7b568e1
# save the config file to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=riscv SHELL=/bin/bash drivers/gpu/drm/amd/amdgpu/
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@...el.com>
All warnings (new ones prefixed by >>):
>> drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:5089: warning: expecting prototype for amdgpu_device_gpu_recover(). Prototype was for amdgpu_device_gpu_recover_imp() instead
vim +5089 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
e6c6338f393b74 Jack Zhang 2021-03-08 5075
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5076 /**
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5077 * amdgpu_device_gpu_recover - reset the asic and recover scheduler
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5078 *
982a820bac1b64 Mauro Carvalho Chehab 2020-10-21 5079 * @adev: amdgpu_device pointer
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5080 * @job: which job trigger hang
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5081 *
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5082 * Attempt to reset the GPU if it has hung (all asics).
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5083 * Attempt to do soft-reset or full-reset and reinitialize Asic
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5084 * Returns 0 for success or an error on failure.
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5085 */
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5086
92fc842d9da15a Andrey Grodzovsky 2021-12-17 5087 int amdgpu_device_gpu_recover_imp(struct amdgpu_device *adev,
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5088 struct amdgpu_job *job)
26bc534094ed45 Andrey Grodzovsky 2018-11-22 @5089 {
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5090 struct list_head device_list, *device_list_handle = NULL;
7dd8c205eaedfa Evan Quan 2020-04-16 5091 bool job_signaled = false;
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5092 struct amdgpu_hive_info *hive = NULL;
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5093 struct amdgpu_device *tmp_adev = NULL;
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5094 int i, r = 0;
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5095 bool need_emergency_restart = false;
3f12acc8d6d4b2 Evan Quan 2020-04-21 5096 bool audio_suspended = false;
e6c6338f393b74 Jack Zhang 2021-03-08 5097 int tmp_vram_lost_counter;
04442bf70debb1 Lijo Lazar 2021-03-16 5098 struct amdgpu_reset_context reset_context;
04442bf70debb1 Lijo Lazar 2021-03-16 5099
04442bf70debb1 Lijo Lazar 2021-03-16 5100 memset(&reset_context, 0, sizeof(reset_context));
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5101
6e3cd2a9a6ac32 Mauro Carvalho Chehab 2020-10-23 5102 /*
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5103 * Special case: RAS triggered and full reset isn't supported
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5104 */
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5105 need_emergency_restart = amdgpu_ras_need_emergency_restart(adev);
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5106
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5107 /*
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5108 * Flush RAM to disk so that after reboot
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5109 * the user can read log and see why the system rebooted.
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5110 */
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5111 if (need_emergency_restart && amdgpu_ras_get_context(adev)->reboot) {
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5112 DRM_WARN("Emergency reboot.");
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5113
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5114 ksys_sync_helper();
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5115 emergency_restart();
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5116 }
d5ea093eebf022 Andrey Grodzovsky 2019-08-22 5117
b823821f2244ad Le Ma 2019-11-27 5118 dev_info(adev->dev, "GPU %s begin!\n",
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5119 need_emergency_restart ? "jobs stop":"reset");
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5120
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5121 /*
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5122 * Here we trylock to avoid chain of resets executing from
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5123 * either trigger by jobs on different adevs in XGMI hive or jobs on
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5124 * different schedulers for same device while this TO handler is running.
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5125 * We always reset all schedulers for device and all devices for XGMI
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5126 * hive so that should take care of them too.
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5127 */
175ac6ec6bd8db Zhigang Luo 2021-11-26 5128 if (!amdgpu_sriov_vf(adev))
d95e8e97e2d522 Dennis Li 2020-08-18 5129 hive = amdgpu_get_xgmi_hive(adev);
53b3f8f40e6cff Dennis Li 2020-08-19 5130 if (hive) {
53b3f8f40e6cff Dennis Li 2020-08-19 5131 if (atomic_cmpxchg(&hive->in_reset, 0, 1) != 0) {
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5132 DRM_INFO("Bailing on TDR for s_job:%llx, hive: %llx as another already in progress",
0b2d2c2eecf27f Andrey Grodzovsky 2019-08-27 5133 job ? job->base.id : -1, hive->hive_id);
d95e8e97e2d522 Dennis Li 2020-08-18 5134 amdgpu_put_xgmi_hive(hive);
ff99849b00fef5 Jingwen Chen 2021-07-20 5135 if (job && job->vm)
91fb309d8294be Horace Chen 2021-01-20 5136 drm_sched_increase_karma(&job->base);
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5137 return 0;
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5138 }
53b3f8f40e6cff Dennis Li 2020-08-19 5139 mutex_lock(&hive->hive_lock);
53b3f8f40e6cff Dennis Li 2020-08-19 5140 }
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5141
04442bf70debb1 Lijo Lazar 2021-03-16 5142 reset_context.method = AMD_RESET_METHOD_NONE;
04442bf70debb1 Lijo Lazar 2021-03-16 5143 reset_context.reset_req_dev = adev;
04442bf70debb1 Lijo Lazar 2021-03-16 5144 reset_context.job = job;
04442bf70debb1 Lijo Lazar 2021-03-16 5145 reset_context.hive = hive;
04442bf70debb1 Lijo Lazar 2021-03-16 5146 clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
04442bf70debb1 Lijo Lazar 2021-03-16 5147
91fb309d8294be Horace Chen 2021-01-20 5148 /*
91fb309d8294be Horace Chen 2021-01-20 5149 * lock the device before we try to operate the linked list
91fb309d8294be Horace Chen 2021-01-20 5150 * if didn't get the device lock, don't touch the linked list since
91fb309d8294be Horace Chen 2021-01-20 5151 * others may iterating it.
91fb309d8294be Horace Chen 2021-01-20 5152 */
91fb309d8294be Horace Chen 2021-01-20 5153 r = amdgpu_device_lock_hive_adev(adev, hive);
91fb309d8294be Horace Chen 2021-01-20 5154 if (r) {
91fb309d8294be Horace Chen 2021-01-20 5155 dev_info(adev->dev, "Bailing on TDR for s_job:%llx, as another already in progress",
91fb309d8294be Horace Chen 2021-01-20 5156 job ? job->base.id : -1);
91fb309d8294be Horace Chen 2021-01-20 5157
91fb309d8294be Horace Chen 2021-01-20 5158 /* even we skipped this reset, still need to set the job to guilty */
ff99849b00fef5 Jingwen Chen 2021-07-20 5159 if (job && job->vm)
91fb309d8294be Horace Chen 2021-01-20 5160 drm_sched_increase_karma(&job->base);
91fb309d8294be Horace Chen 2021-01-20 5161 goto skip_recovery;
91fb309d8294be Horace Chen 2021-01-20 5162 }
91fb309d8294be Horace Chen 2021-01-20 5163
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5164 /*
9e94d22c008585 Evan Quan 2020-04-16 5165 * Build list of devices to reset.
9e94d22c008585 Evan Quan 2020-04-16 5166 * In case we are in XGMI hive mode, resort the device list
9e94d22c008585 Evan Quan 2020-04-16 5167 * to put adev in the 1st position.
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5168 */
9e94d22c008585 Evan Quan 2020-04-16 5169 INIT_LIST_HEAD(&device_list);
175ac6ec6bd8db Zhigang Luo 2021-11-26 5170 if (!amdgpu_sriov_vf(adev) && (adev->gmc.xgmi.num_physical_nodes > 1)) {
655ce9cb13b596 shaoyunl 2021-03-04 5171 list_for_each_entry(tmp_adev, &hive->device_list, gmc.xgmi.head)
655ce9cb13b596 shaoyunl 2021-03-04 5172 list_add_tail(&tmp_adev->reset_list, &device_list);
655ce9cb13b596 shaoyunl 2021-03-04 5173 if (!list_is_first(&adev->reset_list, &device_list))
655ce9cb13b596 shaoyunl 2021-03-04 5174 list_rotate_to_front(&adev->reset_list, &device_list);
655ce9cb13b596 shaoyunl 2021-03-04 5175 device_list_handle = &device_list;
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5176 } else {
655ce9cb13b596 shaoyunl 2021-03-04 5177 list_add_tail(&adev->reset_list, &device_list);
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5178 device_list_handle = &device_list;
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5179 }
26bc534094ed45 Andrey Grodzovsky 2018-11-22 5180
12ffa55da60f83 Andrey Grodzovsky 2019-08-30 5181 /* block all schedulers and reset given job's ring */
655ce9cb13b596 shaoyunl 2021-03-04 5182 list_for_each_entry(tmp_adev, device_list_handle, reset_list) {
3f12acc8d6d4b2 Evan Quan 2020-04-21 5183 /*
3f12acc8d6d4b2 Evan Quan 2020-04-21 5184 * Try to put the audio codec into suspend state
3f12acc8d6d4b2 Evan Quan 2020-04-21 5185 * before gpu reset started.
3f12acc8d6d4b2 Evan Quan 2020-04-21 5186 *
3f12acc8d6d4b2 Evan Quan 2020-04-21 5187 * Due to the power domain of the graphics device
3f12acc8d6d4b2 Evan Quan 2020-04-21 5188 * is shared with AZ power domain. Without this,
3f12acc8d6d4b2 Evan Quan 2020-04-21 5189 * we may change the audio hardware from behind
3f12acc8d6d4b2 Evan Quan 2020-04-21 5190 * the audio driver's back. That will trigger
3f12acc8d6d4b2 Evan Quan 2020-04-21 5191 * some audio codec errors.
3f12acc8d6d4b2 Evan Quan 2020-04-21 5192 */
3f12acc8d6d4b2 Evan Quan 2020-04-21 5193 if (!amdgpu_device_suspend_display_audio(tmp_adev))
3f12acc8d6d4b2 Evan Quan 2020-04-21 5194 audio_suspended = true;
3f12acc8d6d4b2 Evan Quan 2020-04-21 5195
9e94d22c008585 Evan Quan 2020-04-16 5196 amdgpu_ras_set_error_query_ready(tmp_adev, false);
9e94d22c008585 Evan Quan 2020-04-16 5197
52fb44cf30fc6b Evan Quan 2020-04-16 5198 cancel_delayed_work_sync(&tmp_adev->delayed_init_work);
52fb44cf30fc6b Evan Quan 2020-04-16 5199
428890a3fec131 shaoyunl 2021-11-29 5200 if (!amdgpu_sriov_vf(tmp_adev))
9e94d22c008585 Evan Quan 2020-04-16 5201 amdgpu_amdkfd_pre_reset(tmp_adev);
9e94d22c008585 Evan Quan 2020-04-16 5202
fdafb3597a2cc4 Evan Quan 2019-06-26 5203 /*
fdafb3597a2cc4 Evan Quan 2019-06-26 5204 * Mark these ASICs to be reseted as untracked first
fdafb3597a2cc4 Evan Quan 2019-06-26 5205 * And add them back after reset completed
fdafb3597a2cc4 Evan Quan 2019-06-26 5206 */
fdafb3597a2cc4 Evan Quan 2019-06-26 5207 amdgpu_unregister_gpu_instance(tmp_adev);
fdafb3597a2cc4 Evan Quan 2019-06-26 5208
087451f372bf76 Evan Quan 2021-10-19 5209 drm_fb_helper_set_suspend_unlocked(adev_to_drm(adev)->fb_helper, true);
565d1941557756 Evan Quan 2020-03-11 5210
f1c1314be42971 xinhui pan 2019-07-04 5211 /* disable ras on ALL IPs */
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5212 if (!need_emergency_restart &&
b823821f2244ad Le Ma 2019-11-27 5213 amdgpu_device_ip_need_full_reset(tmp_adev))
f1c1314be42971 xinhui pan 2019-07-04 5214 amdgpu_ras_suspend(tmp_adev);
f1c1314be42971 xinhui pan 2019-07-04 5215
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5216 for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5217 struct amdgpu_ring *ring = tmp_adev->rings[i];
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5218
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5219 if (!ring || !ring->sched.thread)
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5220 continue;
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5221
0b2d2c2eecf27f Andrey Grodzovsky 2019-08-27 5222 drm_sched_stop(&ring->sched, job ? &job->base : NULL);
7c6e68c777f109 Andrey Grodzovsky 2019-09-13 5223
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5224 if (need_emergency_restart)
7c6e68c777f109 Andrey Grodzovsky 2019-09-13 5225 amdgpu_job_stop_all_jobs_on_sched(&ring->sched);
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5226 }
8f8c80f4300967 Jingwen Chen 2021-02-25 5227 atomic_inc(&tmp_adev->gpu_reset_counter);
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5228 }
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5229
bb5c7235eaafb4 Wenhui Sheng 2020-07-13 5230 if (need_emergency_restart)
7c6e68c777f109 Andrey Grodzovsky 2019-09-13 5231 goto skip_sched_resume;
7c6e68c777f109 Andrey Grodzovsky 2019-09-13 5232
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5233 /*
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5234 * Must check guilty signal here since after this point all old
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5235 * HW fences are force signaled.
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5236 *
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5237 * job->base holds a reference to parent fence
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5238 */
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5239 if (job && job->base.s_fence->parent &&
7dd8c205eaedfa Evan Quan 2020-04-16 5240 dma_fence_is_signaled(job->base.s_fence->parent)) {
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5241 job_signaled = true;
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5242 dev_info(adev->dev, "Guilty job already signaled, skipping HW reset");
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5243 goto skip_hw_reset;
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5244 }
1d721ed679db18 Andrey Grodzovsky 2019-04-18 5245
:::::: The code at line 5089 was first introduced by commit
:::::: 26bc534094ed45fdedef6b4ce8b96030340c5ce7 drm/amdgpu: Refactor GPU reset for XGMI hive case
:::::: TO: Andrey Grodzovsky <andrey.grodzovsky@....com>
:::::: CC: Alex Deucher <alexander.deucher@....com>
---
0-DAY CI Kernel Test Service
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
Powered by blists - more mailing lists