From 3eb55e6f753a379e293395de8d5f3be28351a7f8 Mon Sep 17 00:00:00 2001 From: Tina Zhang Date: Fri, 21 Feb 2020 10:32:34 +0800 Subject: [PATCH 01/11] drm/i915/gvt: Separate display reset from ALL_ENGINES reset ALL_ENGINES reset doesn't clobber display with the current gvt-g supported platforms. Thus ALL_ENGINES reset shouldn't reset the display engine registers emulated by gvt-g. This fixes guest warning like [ 14.622026] [drm] Initialized i915 1.6.0 20200114 for 0000:00:03.0 on minor 0 [ 14.967917] fbcon: i915drmfb (fb0) is primary device [ 25.100188] [drm:drm_atomic_helper_wait_for_dependencies [drm_kms_helper]] E RROR [CRTC:51:pipe A] flip_done timed out [ 25.100860] -----------[ cut here ]----------- [ 25.100861] pll on state mismatch (expected 0, found 1) [ 25.101024] WARNING: CPU: 1 PID: 30 at drivers/gpu/drm/i915/display/intel_dis play.c:14382 verify_single_dpll_state.isra.115+0x28f/0x320 [i915] [ 25.101025] Modules linked in: intel_rapl_msr intel_rapl_common kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel i915 aesni_intel cr ypto_simd cryptd glue_helper cec rc_core video drm_kms_helper joydev drm input_l eds i2c_algo_bit serio_raw fb_sys_fops syscopyarea sysfillrect sysimgblt mac_hid qemu_fw_cfg sch_fq_codel parport_pc ppdev lp parport ip_tables x_tables autofs4 e1000 psmouse i2c_piix4 pata_acpi floppy [ 25.101052] CPU: 1 PID: 30 Comm: kworker/u4:1 Not tainted 5.5.0+ #1 [ 25.101053] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1 .12.1-0-ga5cab58 04/01/2014 [ 25.101055] Workqueue: events_unbound async_run_entry_fn [ 25.101092] RIP: 0010:verify_single_dpll_state.isra.115+0x28f/0x320 [i915] [ 25.101093] Code: e0 d9 ff e9 a3 fe ff ff 80 3d e9 c2 11 00 00 44 89 f6 48 c7 c7 c0 9d 88 c0 75 3b e8 eb df d9 ff e9 c7 fe ff ff e8 d1 e0 ae c4 <0f> 0b e9 7a fe ff ff 80 3d c0 c2 11 00 00 8d 71 41 89 c2 48 c7 c7 [ 25.101093] RSP: 0018:ffffb1de80107878 EFLAGS: 00010286 [ 25.101094] RAX: 0000000000000000 RBX: ffffb1de80107884 RCX: 0000000000000007 [ 25.101095] RDX: 0000000000000000 RSI: 0000000000000002 RDI: ffff94fdfdd19740 [ 25.101095] RBP: ffffb1de80107938 R08: 0000000d6bfdc7b4 R09: 000000000000002b [ 25.101096] R10: ffff94fdf82dc000 R11: 0000000000000225 R12: 00000000000001f8 [ 25.101096] R13: ffff94fdb3ca6a90 R14: ffff94fdb3ca0000 R15: 0000000000000000 [ 25.101097] FS: 0000000000000000(0000) GS:ffff94fdfdd00000(0000) knlGS:00000 00000000000 [ 25.101098] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 25.101098] CR2: 00007fbc3e2be9c8 CR3: 000000003339a003 CR4: 0000000000360ee0 [ 25.101101] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 25.101101] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 25.101102] Call Trace: [ 25.101139] intel_atomic_commit_tail+0xde4/0x1520 [i915] [ 25.101141] ? flush_workqueue_prep_pwqs+0xfa/0x130 [ 25.101142] ? flush_workqueue+0x198/0x3c0 [ 25.101174] intel_atomic_commit+0x2ad/0x320 [i915] [ 25.101209] drm_atomic_commit+0x4a/0x50 [drm] [ 25.101220] drm_client_modeset_commit_atomic+0x1c4/0x200 [drm] [ 25.101231] drm_client_modeset_commit_force+0x47/0x170 [drm] [ 25.101250] drm_fb_helper_restore_fbdev_mode_unlocked+0x4e/0xa0 [drm_kms_hel per] [ 25.101255] drm_fb_helper_set_par+0x2d/0x60 [drm_kms_helper] [ 25.101287] intel_fbdev_set_par+0x1a/0x40 [i915] [ 25.101289] ? con_is_visible+0x2e/0x60 [ 25.101290] fbcon_init+0x378/0x600 [ 25.101292] visual_init+0xd5/0x130 [ 25.101296] do_bind_con_driver+0x217/0x430 [ 25.101297] do_take_over_console+0x7d/0x1b0 [ 25.101298] do_fbcon_takeover+0x5c/0xb0 [ 25.101299] fbcon_fb_registered+0x199/0x1a0 [ 25.101301] register_framebuffer+0x22c/0x330 [ 25.101306] __drm_fb_helper_initial_config_and_unlock+0x31a/0x520 [drm_kms_h elper] [ 25.101311] drm_fb_helper_initial_config+0x35/0x40 [drm_kms_helper] [ 25.101341] intel_fbdev_initial_config+0x18/0x30 [i915] [ 25.101342] async_run_entry_fn+0x3c/0x150 [ 25.101343] process_one_work+0x1fd/0x3f0 [ 25.101344] worker_thread+0x34/0x410 [ 25.101346] kthread+0x121/0x140 [ 25.101346] ? process_one_work+0x3f0/0x3f0 [ 25.101347] ? kthread_park+0x90/0x90 [ 25.101350] ret_from_fork+0x35/0x40 [ 25.101351] --[ end trace b5b47d44cd998ba1 ]-- Fixes: 6294b61ba769 ("drm/i915/gvt: add missing display part reset for vGPU reset") Signed-off-by: Tina Zhang Reviewed-by: Zhenyu Wang Signed-off-by: Zhenyu Wang Link: http://patchwork.freedesktop.org/patch/msgid/20200221023234.28635-1-tina.zhang@intel.com --- drivers/gpu/drm/i915/gvt/vgpu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c index 85bd9bf4f6ee..487af6ea9972 100644 --- a/drivers/gpu/drm/i915/gvt/vgpu.c +++ b/drivers/gpu/drm/i915/gvt/vgpu.c @@ -560,9 +560,9 @@ void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr, intel_vgpu_reset_mmio(vgpu, dmlr); populate_pvinfo_page(vgpu); - intel_vgpu_reset_display(vgpu); if (dmlr) { + intel_vgpu_reset_display(vgpu); intel_vgpu_reset_cfg_space(vgpu); /* only reset the failsafe mode when dmlr reset */ vgpu->failsafe = false; From b549c252b1292aea959cd9b83537fcb9384a6112 Mon Sep 17 00:00:00 2001 From: Tina Zhang Date: Tue, 25 Feb 2020 13:35:27 +0800 Subject: [PATCH 02/11] drm/i915/gvt: Fix orphan vgpu dmabuf_objs' lifetime Deleting dmabuf item's list head after releasing its container can lead to KASAN-reported issue: BUG: KASAN: use-after-free in __list_del_entry_valid+0x15/0xf0 Read of size 8 at addr ffff88818a4598a8 by task kworker/u8:3/13119 So fix this issue by puting deleting dmabuf_objs ahead of releasing its container. Fixes: dfb6ae4e14bd6 ("drm/i915/gvt: Handle orphan dmabuf_objs") Signed-off-by: Tina Zhang Reviewed-by: Zhenyu Wang Signed-off-by: Zhenyu Wang Link: http://patchwork.freedesktop.org/patch/msgid/20200225053527.8336-2-tina.zhang@intel.com --- drivers/gpu/drm/i915/gvt/dmabuf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gvt/dmabuf.c b/drivers/gpu/drm/i915/gvt/dmabuf.c index 2477a1e5a166..ae139f0877ae 100644 --- a/drivers/gpu/drm/i915/gvt/dmabuf.c +++ b/drivers/gpu/drm/i915/gvt/dmabuf.c @@ -151,12 +151,12 @@ static void dmabuf_gem_object_free(struct kref *kref) dmabuf_obj = container_of(pos, struct intel_vgpu_dmabuf_obj, list); if (dmabuf_obj == obj) { + list_del(pos); intel_gvt_hypervisor_put_vfio_device(vgpu); idr_remove(&vgpu->object_idr, dmabuf_obj->dmabuf_id); kfree(dmabuf_obj->info); kfree(dmabuf_obj); - list_del(pos); break; } } From 4829f89855f1d3a3d8014e74cceab51b421503db Mon Sep 17 00:00:00 2001 From: Monk Liu Date: Sat, 8 Feb 2020 19:01:21 +0800 Subject: [PATCH 03/11] drm/amdgpu: fix memory leak during TDR test(v2) fix system memory leak v2: fix coding style Signed-off-by: Monk Liu Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/powerplay/smu_v11_0.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c index b06c057a9002..c9e5ce135fd4 100644 --- a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c +++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c @@ -978,8 +978,12 @@ int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu) struct smu_11_0_max_sustainable_clocks *max_sustainable_clocks; int ret = 0; - max_sustainable_clocks = kzalloc(sizeof(struct smu_11_0_max_sustainable_clocks), + if (!smu->smu_table.max_sustainable_clocks) + max_sustainable_clocks = kzalloc(sizeof(struct smu_11_0_max_sustainable_clocks), GFP_KERNEL); + else + max_sustainable_clocks = smu->smu_table.max_sustainable_clocks; + smu->smu_table.max_sustainable_clocks = (void *)max_sustainable_clocks; max_sustainable_clocks->uclock = smu->smu_table.boot_values.uclk / 100; From a3ed353cf8015ba84a0407a5dc3ffee038166ab0 Mon Sep 17 00:00:00 2001 From: Shirish S Date: Mon, 27 Jan 2020 16:35:24 +0530 Subject: [PATCH 04/11] amdgpu/gmc_v9: save/restore sdpif regs during S3 fixes S3 issue with IOMMU + S/G enabled @ 64M VRAM. Suggested-by: Alex Deucher Signed-off-by: Shirish S Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h | 1 + drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 37 ++++++++++++++++++- .../include/asic_reg/dce/dce_12_0_offset.h | 2 + 3 files changed, 39 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h index d3c27a3c43f6..7546da0cc70c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h @@ -195,6 +195,7 @@ struct amdgpu_gmc { uint32_t srbm_soft_reset; bool prt_warning; uint64_t stolen_size; + uint32_t sdpif_register; /* apertures */ u64 shared_aperture_start; u64 shared_aperture_end; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 90216abf14a4..cc0c273a86f9 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -1271,6 +1271,19 @@ static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev) } } +/** + * gmc_v9_0_restore_registers - restores regs + * + * @adev: amdgpu_device pointer + * + * This restores register values, saved at suspend. + */ +static void gmc_v9_0_restore_registers(struct amdgpu_device *adev) +{ + if (adev->asic_type == CHIP_RAVEN) + WREG32(mmDCHUBBUB_SDPIF_MMIO_CNTRL_0, adev->gmc.sdpif_register); +} + /** * gmc_v9_0_gart_enable - gart enable * @@ -1376,6 +1389,20 @@ static int gmc_v9_0_hw_init(void *handle) return r; } +/** + * gmc_v9_0_save_registers - saves regs + * + * @adev: amdgpu_device pointer + * + * This saves potential register values that should be + * restored upon resume + */ +static void gmc_v9_0_save_registers(struct amdgpu_device *adev) +{ + if (adev->asic_type == CHIP_RAVEN) + adev->gmc.sdpif_register = RREG32(mmDCHUBBUB_SDPIF_MMIO_CNTRL_0); +} + /** * gmc_v9_0_gart_disable - gart disable * @@ -1412,9 +1439,16 @@ static int gmc_v9_0_hw_fini(void *handle) static int gmc_v9_0_suspend(void *handle) { + int r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; - return gmc_v9_0_hw_fini(adev); + r = gmc_v9_0_hw_fini(adev); + if (r) + return r; + + gmc_v9_0_save_registers(adev); + + return 0; } static int gmc_v9_0_resume(void *handle) @@ -1422,6 +1456,7 @@ static int gmc_v9_0_resume(void *handle) int r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + gmc_v9_0_restore_registers(adev); r = gmc_v9_0_hw_init(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h index b6f74bf4af02..27bb8c1ab858 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h @@ -7376,6 +7376,8 @@ #define mmCRTC4_CRTC_DRR_CONTROL 0x0f3e #define mmCRTC4_CRTC_DRR_CONTROL_BASE_IDX 2 +#define mmDCHUBBUB_SDPIF_MMIO_CNTRL_0 0x395d +#define mmDCHUBBUB_SDPIF_MMIO_CNTRL_0_BASE_IDX 2 // addressBlock: dce_dc_fmt4_dispdec // base address: 0x2000 From b5dacc8fb52c690e2cdf7df3ae36bd1cf20e63dd Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Fri, 21 Feb 2020 12:54:14 +0200 Subject: [PATCH 05/11] drm/i915: fix header test with GCOV $(CC) with $(CFLAGS_GCOV) assumes the output filename with .gcno suffix appended is writable. This is not the case when the output filename is /dev/null: HDRTEST drivers/gpu/drm/i915/display/intel_frontbuffer.h /dev/null:1:0: error: cannot open /dev/null.gcno HDRTEST drivers/gpu/drm/i915/display/intel_ddi.h /dev/null:1:0: error: cannot open /dev/null.gcno make[5]: *** [../drivers/gpu/drm/i915/Makefile:307: drivers/gpu/drm/i915/display/intel_ddi.hdrtest] Error 1 make[5]: *** Waiting for unfinished jobs.... make[5]: *** [../drivers/gpu/drm/i915/Makefile:307: drivers/gpu/drm/i915/display/intel_frontbuffer.hdrtest] Error 1 Filter out $(CFLAGS_GVOC) from the header test $(c_flags) as they don't make sense here anyway. References: http://lore.kernel.org/r/d8112767-4089-4c58-d7d3-2ce03139858a@infradead.org Reported-by: Randy Dunlap Fixes: c6d4a099a240 ("drm/i915: reimplement header test feature") Cc: Masahiro Yamada Acked-by: Randy Dunlap Signed-off-by: Jani Nikula Link: https://patchwork.freedesktop.org/patch/msgid/20200221105414.14358-1-jani.nikula@intel.com (cherry picked from commit 408c1b3253dab93da175690dc0e21dd8bccf3371) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index b8c5f8934dbd..a1f2411aa21b 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -294,7 +294,7 @@ extra-$(CONFIG_DRM_I915_WERROR) += \ $(shell cd $(srctree)/$(src) && find * -name '*.h'))) quiet_cmd_hdrtest = HDRTEST $(patsubst %.hdrtest,%.h,$@) - cmd_hdrtest = $(CC) $(c_flags) -S -o /dev/null -x c /dev/null -include $<; touch $@ + cmd_hdrtest = $(CC) $(filter-out $(CFLAGS_GCOV), $(c_flags)) -S -o /dev/null -x c /dev/null -include $<; touch $@ $(obj)/%.hdrtest: $(src)/%.h FORCE $(call if_changed_dep,hdrtest) From eee18939e5767dbe3a98b3ea172e7fd7ba7d403c Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 24 Feb 2020 10:11:20 +0000 Subject: [PATCH 06/11] drm/i915/gtt: Downgrade gen7 (ivb, byt, hsw) back to aliasing-ppgtt Full-ppgtt on gen7 is proving to be highly unstable and not robust. Closes: https://gitlab.freedesktop.org/drm/intel/issues/694 Fixes: 3cd6e8860ecd ("drm/i915/gen7: Re-enable full-ppgtt for ivb & hsw") Signed-off-by: Chris Wilson Cc: Joonas Lahtinen Cc: Rodrigo Vivi Cc: Jani Nikula Cc: Dave Airlie Acked-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20200224101120.4024481-1-chris@chris-wilson.co.uk (cherry picked from commit 4fbe112a569526e46fa2accb5763c069f78cb431) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_pci.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c index 83f01401b8b5..f631f6d21127 100644 --- a/drivers/gpu/drm/i915/i915_pci.c +++ b/drivers/gpu/drm/i915/i915_pci.c @@ -437,7 +437,7 @@ static const struct intel_device_info snb_m_gt2_info = { .has_rc6 = 1, \ .has_rc6p = 1, \ .has_rps = true, \ - .ppgtt_type = INTEL_PPGTT_FULL, \ + .ppgtt_type = INTEL_PPGTT_ALIASING, \ .ppgtt_size = 31, \ IVB_PIPE_OFFSETS, \ IVB_CURSOR_OFFSETS, \ @@ -494,7 +494,7 @@ static const struct intel_device_info vlv_info = { .has_rps = true, .display.has_gmch = 1, .display.has_hotplug = 1, - .ppgtt_type = INTEL_PPGTT_FULL, + .ppgtt_type = INTEL_PPGTT_ALIASING, .ppgtt_size = 31, .has_snoop = true, .has_coherent_ggtt = false, From 19ee5e8da6129d8d828201a12264ab3d09153ec4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20Winiarski?= Date: Wed, 19 Feb 2020 17:18:21 +0100 Subject: [PATCH 07/11] drm/i915/pmu: Avoid using globals for CPU hotplug state MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Attempting to bind / unbind module from devices where we have both integrated and discreete GPU handled by i915 can lead to leaks and warnings from cpuhp: Error: Removing state XXX which has instances left. Let's move the state to i915_pmu. Fixes: 05488673a4d4 ("drm/i915/pmu: Support multiple GPUs") Signed-off-by: Michał Winiarski Cc: Chris Wilson Cc: Michal Wajdeczko Cc: Tvrtko Ursulin Reviewed-by: Chris Wilson Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20200219161822.24592-1-michal.winiarski@intel.com (cherry picked from commit f5a179d4687d4e7bfadd7cbda7ee5d0bad76761f) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_pmu.c | 18 +++++++++--------- drivers/gpu/drm/i915/i915_pmu.h | 7 +++++-- 2 files changed, 14 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c index ec0299490dd4..84301004d5c0 100644 --- a/drivers/gpu/drm/i915/i915_pmu.c +++ b/drivers/gpu/drm/i915/i915_pmu.c @@ -1042,7 +1042,7 @@ static void free_event_attributes(struct i915_pmu *pmu) static int i915_pmu_cpu_online(unsigned int cpu, struct hlist_node *node) { - struct i915_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), node); + struct i915_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), cpuhp.node); GEM_BUG_ON(!pmu->base.event_init); @@ -1055,7 +1055,7 @@ static int i915_pmu_cpu_online(unsigned int cpu, struct hlist_node *node) static int i915_pmu_cpu_offline(unsigned int cpu, struct hlist_node *node) { - struct i915_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), node); + struct i915_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), cpuhp.node); unsigned int target; GEM_BUG_ON(!pmu->base.event_init); @@ -1072,8 +1072,6 @@ static int i915_pmu_cpu_offline(unsigned int cpu, struct hlist_node *node) return 0; } -static enum cpuhp_state cpuhp_slot = CPUHP_INVALID; - static int i915_pmu_register_cpuhp_state(struct i915_pmu *pmu) { enum cpuhp_state slot; @@ -1087,21 +1085,22 @@ static int i915_pmu_register_cpuhp_state(struct i915_pmu *pmu) return ret; slot = ret; - ret = cpuhp_state_add_instance(slot, &pmu->node); + ret = cpuhp_state_add_instance(slot, &pmu->cpuhp.node); if (ret) { cpuhp_remove_multi_state(slot); return ret; } - cpuhp_slot = slot; + pmu->cpuhp.slot = slot; return 0; } static void i915_pmu_unregister_cpuhp_state(struct i915_pmu *pmu) { - WARN_ON(cpuhp_slot == CPUHP_INVALID); - WARN_ON(cpuhp_state_remove_instance(cpuhp_slot, &pmu->node)); - cpuhp_remove_multi_state(cpuhp_slot); + WARN_ON(pmu->cpuhp.slot == CPUHP_INVALID); + WARN_ON(cpuhp_state_remove_instance(pmu->cpuhp.slot, &pmu->cpuhp.node)); + cpuhp_remove_multi_state(pmu->cpuhp.slot); + pmu->cpuhp.slot = CPUHP_INVALID; } static bool is_igp(struct drm_i915_private *i915) @@ -1128,6 +1127,7 @@ void i915_pmu_register(struct drm_i915_private *i915) spin_lock_init(&pmu->lock); hrtimer_init(&pmu->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); pmu->timer.function = i915_sample; + pmu->cpuhp.slot = CPUHP_INVALID; if (!is_igp(i915)) { pmu->name = kasprintf(GFP_KERNEL, diff --git a/drivers/gpu/drm/i915/i915_pmu.h b/drivers/gpu/drm/i915/i915_pmu.h index 6c1647c5daf2..207058391cec 100644 --- a/drivers/gpu/drm/i915/i915_pmu.h +++ b/drivers/gpu/drm/i915/i915_pmu.h @@ -39,9 +39,12 @@ struct i915_pmu_sample { struct i915_pmu { /** - * @node: List node for CPU hotplug handling. + * @cpuhp: Struct used for CPU hotplug handling. */ - struct hlist_node node; + struct { + struct hlist_node node; + enum cpuhp_state slot; + } cpuhp; /** * @base: PMU base. */ From 2de0147d77168d6a227c00eb9c5a49374e1582a3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20Winiarski?= Date: Wed, 19 Feb 2020 17:18:22 +0100 Subject: [PATCH 08/11] drm/i915/pmu: Avoid using globals for PMU events MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Attempting to bind / unbind module from devices where we have both integrated and discreete GPU handled by i915, will cause us to try and double free the global state, hitting null ptr deref in free_event_attributes. Let's move it to i915_pmu. Fixes: 05488673a4d4 ("drm/i915/pmu: Support multiple GPUs") Signed-off-by: Michał Winiarski Cc: Chris Wilson Cc: Michal Wajdeczko Cc: Tvrtko Ursulin Reviewed-by: Chris Wilson Signed-off-by: Chris Wilson Link: https://patchwork.freedesktop.org/patch/msgid/20200219161822.24592-2-michal.winiarski@intel.com (cherry picked from commit 46129dc10f47c5c2b51c93a82b7b2aca46574ae0) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_pmu.c | 41 ++++++++++++++++++--------------- drivers/gpu/drm/i915/i915_pmu.h | 4 ++++ 2 files changed, 26 insertions(+), 19 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c index 84301004d5c0..aa729d04abe2 100644 --- a/drivers/gpu/drm/i915/i915_pmu.c +++ b/drivers/gpu/drm/i915/i915_pmu.c @@ -822,11 +822,6 @@ static ssize_t i915_pmu_event_show(struct device *dev, return sprintf(buf, "config=0x%lx\n", eattr->val); } -static struct attribute_group i915_pmu_events_attr_group = { - .name = "events", - /* Patch in attrs at runtime. */ -}; - static ssize_t i915_pmu_get_attr_cpumask(struct device *dev, struct device_attribute *attr, @@ -846,13 +841,6 @@ static const struct attribute_group i915_pmu_cpumask_attr_group = { .attrs = i915_cpumask_attrs, }; -static const struct attribute_group *i915_pmu_attr_groups[] = { - &i915_pmu_format_attr_group, - &i915_pmu_events_attr_group, - &i915_pmu_cpumask_attr_group, - NULL -}; - #define __event(__config, __name, __unit) \ { \ .config = (__config), \ @@ -1026,16 +1014,16 @@ err:; static void free_event_attributes(struct i915_pmu *pmu) { - struct attribute **attr_iter = i915_pmu_events_attr_group.attrs; + struct attribute **attr_iter = pmu->events_attr_group.attrs; for (; *attr_iter; attr_iter++) kfree((*attr_iter)->name); - kfree(i915_pmu_events_attr_group.attrs); + kfree(pmu->events_attr_group.attrs); kfree(pmu->i915_attr); kfree(pmu->pmu_attr); - i915_pmu_events_attr_group.attrs = NULL; + pmu->events_attr_group.attrs = NULL; pmu->i915_attr = NULL; pmu->pmu_attr = NULL; } @@ -1117,6 +1105,13 @@ static bool is_igp(struct drm_i915_private *i915) void i915_pmu_register(struct drm_i915_private *i915) { struct i915_pmu *pmu = &i915->pmu; + const struct attribute_group *attr_groups[] = { + &i915_pmu_format_attr_group, + &pmu->events_attr_group, + &i915_pmu_cpumask_attr_group, + NULL + }; + int ret = -ENOMEM; if (INTEL_GEN(i915) <= 2) { @@ -1143,11 +1138,16 @@ void i915_pmu_register(struct drm_i915_private *i915) if (!pmu->name) goto err; - i915_pmu_events_attr_group.attrs = create_event_attributes(pmu); - if (!i915_pmu_events_attr_group.attrs) + pmu->events_attr_group.name = "events"; + pmu->events_attr_group.attrs = create_event_attributes(pmu); + if (!pmu->events_attr_group.attrs) goto err_name; - pmu->base.attr_groups = i915_pmu_attr_groups; + pmu->base.attr_groups = kmemdup(attr_groups, sizeof(attr_groups), + GFP_KERNEL); + if (!pmu->base.attr_groups) + goto err_attr; + pmu->base.task_ctx_nr = perf_invalid_context; pmu->base.event_init = i915_pmu_event_init; pmu->base.add = i915_pmu_event_add; @@ -1159,7 +1159,7 @@ void i915_pmu_register(struct drm_i915_private *i915) ret = perf_pmu_register(&pmu->base, pmu->name, -1); if (ret) - goto err_attr; + goto err_groups; ret = i915_pmu_register_cpuhp_state(pmu); if (ret) @@ -1169,6 +1169,8 @@ void i915_pmu_register(struct drm_i915_private *i915) err_unreg: perf_pmu_unregister(&pmu->base); +err_groups: + kfree(pmu->base.attr_groups); err_attr: pmu->base.event_init = NULL; free_event_attributes(pmu); @@ -1194,6 +1196,7 @@ void i915_pmu_unregister(struct drm_i915_private *i915) perf_pmu_unregister(&pmu->base); pmu->base.event_init = NULL; + kfree(pmu->base.attr_groups); if (!is_igp(i915)) kfree(pmu->name); free_event_attributes(pmu); diff --git a/drivers/gpu/drm/i915/i915_pmu.h b/drivers/gpu/drm/i915/i915_pmu.h index 207058391cec..f1d6cad0d7d5 100644 --- a/drivers/gpu/drm/i915/i915_pmu.h +++ b/drivers/gpu/drm/i915/i915_pmu.h @@ -107,6 +107,10 @@ struct i915_pmu { * @sleep_last: Last time GT parked for RC6 estimation. */ ktime_t sleep_last; + /** + * @events_attr_group: Device events attribute group. + */ + struct attribute_group events_attr_group; /** * @i915_attr: Memory block holding device attributes. */ From 238734262142075056653b4de091458e0ca858f2 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 21 Feb 2020 22:18:18 +0000 Subject: [PATCH 09/11] drm/i915: Avoid recursing onto active vma from the shrinker We mark the vma as active while binding it in order to protect outselves from being shrunk under mempressure. This only works if we are strict in not attempting to shrink active objects. <6> [472.618968] Workqueue: events_unbound fence_work [i915] <4> [472.618970] Call Trace: <4> [472.618974] ? __schedule+0x2e5/0x810 <4> [472.618978] schedule+0x37/0xe0 <4> [472.618982] schedule_preempt_disabled+0xf/0x20 <4> [472.618984] __mutex_lock+0x281/0x9c0 <4> [472.618987] ? mark_held_locks+0x49/0x70 <4> [472.618989] ? _raw_spin_unlock_irqrestore+0x47/0x60 <4> [472.619038] ? i915_vma_unbind+0xae/0x110 [i915] <4> [472.619084] ? i915_vma_unbind+0xae/0x110 [i915] <4> [472.619122] i915_vma_unbind+0xae/0x110 [i915] <4> [472.619165] i915_gem_object_unbind+0x1dc/0x400 [i915] <4> [472.619208] i915_gem_shrink+0x328/0x660 [i915] <4> [472.619250] ? i915_gem_shrink_all+0x38/0x60 [i915] <4> [472.619282] i915_gem_shrink_all+0x38/0x60 [i915] <4> [472.619325] vm_alloc_page.constprop.25+0x1aa/0x240 [i915] <4> [472.619330] ? rcu_read_lock_sched_held+0x4d/0x80 <4> [472.619363] ? __alloc_pd+0xb/0x30 [i915] <4> [472.619366] ? module_assert_mutex_or_preempt+0xf/0x30 <4> [472.619368] ? __module_address+0x23/0xe0 <4> [472.619371] ? is_module_address+0x26/0x40 <4> [472.619374] ? static_obj+0x34/0x50 <4> [472.619376] ? lockdep_init_map+0x4d/0x1e0 <4> [472.619407] setup_page_dma+0xd/0x90 [i915] <4> [472.619437] alloc_pd+0x29/0x50 [i915] <4> [472.619470] __gen8_ppgtt_alloc+0x443/0x6b0 [i915] <4> [472.619503] gen8_ppgtt_alloc+0xd7/0x300 [i915] <4> [472.619535] ppgtt_bind_vma+0x2a/0xe0 [i915] <4> [472.619577] __vma_bind+0x26/0x40 [i915] <4> [472.619611] fence_work+0x1c/0x90 [i915] <4> [472.619617] process_one_work+0x26a/0x620 Fixes: 2850748ef876 ("drm/i915: Pull i915_vma_pin under the vm->mutex") Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Reviewed-by: Tvrtko Ursulin Link: https://patchwork.freedesktop.org/patch/msgid/20200221221818.2861432-1-chris@chris-wilson.co.uk (cherry picked from commit 6f24e41022f28061368776ea1514db0a6e67a9b1) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/gem/i915_gem_shrinker.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c index f7e4b39c734f..59b387ade49c 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c @@ -256,8 +256,7 @@ unsigned long i915_gem_shrink_all(struct drm_i915_private *i915) with_intel_runtime_pm(&i915->runtime_pm, wakeref) { freed = i915_gem_shrink(i915, -1UL, NULL, I915_SHRINK_BOUND | - I915_SHRINK_UNBOUND | - I915_SHRINK_ACTIVE); + I915_SHRINK_UNBOUND); } return freed; @@ -336,7 +335,6 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) freed_pages = 0; with_intel_runtime_pm(&i915->runtime_pm, wakeref) freed_pages += i915_gem_shrink(i915, -1UL, NULL, - I915_SHRINK_ACTIVE | I915_SHRINK_BOUND | I915_SHRINK_UNBOUND | I915_SHRINK_WRITEBACK); From 8a3bddf67ce88b96531fb22c5a75d7f4dc41d155 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Sat, 22 Feb 2020 18:54:31 +0100 Subject: [PATCH 10/11] drm/amdgpu: Drop DRIVER_USE_AGP MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This doesn't do anything except auto-init drm_agp support when you call drm_get_pci_dev(). Which amdgpu stopped doing with commit b58c11314a1706bf094c489ef5cb28f76478c704 Author: Alex Deucher Date: Fri Jun 2 17:16:31 2017 -0400 drm/amdgpu: drop deprecated drm_get_pci_dev and drm_put_dev No idea whether this was intentional or accidental breakage, but I guess anyone who manages to boot a this modern gpu behind an agp bridge deserves a price. A price I never expect anyone to ever collect :-) Cc: Alex Deucher Cc: "Christian König" Cc: Hawking Zhang Cc: Xiaojie Yuan Cc: Evan Quan Cc: "Tianci.Yin" Cc: "Marek Olšák" Cc: Hans de Goede Reviewed-by: Emil Velikov Reviewed-by: Alex Deucher Signed-off-by: Daniel Vetter Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 94e2fd758e01..42f4febe24c6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -1389,7 +1389,7 @@ amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe, static struct drm_driver kms_driver = { .driver_features = - DRIVER_USE_AGP | DRIVER_ATOMIC | + DRIVER_ATOMIC | DRIVER_GEM | DRIVER_RENDER | DRIVER_MODESET | DRIVER_SYNCOBJ | DRIVER_SYNCOBJ_TIMELINE, From eb12c957735b582607e5842a06d1f4c62e185c1d Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Sat, 22 Feb 2020 18:54:32 +0100 Subject: [PATCH 11/11] drm/radeon: Inline drm_get_pci_dev MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It's the last user, and more importantly, it's the last non-legacy user of anything in drm_pci.c. The only tricky bit is the agp initialization. But a close look shows that radeon does not use the drm_agp midlayer (the main use of that is drm_bufs for legacy drivers), and instead could use the agp subsystem directly (like nouveau does already). Hence we can just pull this in too. A further step would be to entirely drop the use of drm_device->agp, but feels like too much churn just for this patch. Signed-off-by: Daniel Vetter Cc: Alex Deucher Cc: "Christian König" Cc: "David (ChunMing) Zhou" Cc: amd-gfx@lists.freedesktop.org Reviewed-by: Alex Deucher Reviewed-by: Emil Velikov Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/radeon/radeon_drv.c | 43 +++++++++++++++++++++++++++-- drivers/gpu/drm/radeon/radeon_kms.c | 6 ++++ 2 files changed, 47 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index fd74e2611185..8696af1ee14d 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -37,6 +37,7 @@ #include #include +#include #include #include #include @@ -325,6 +326,7 @@ static int radeon_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent) { unsigned long flags = 0; + struct drm_device *dev; int ret; if (!ent) @@ -365,7 +367,44 @@ static int radeon_pci_probe(struct pci_dev *pdev, if (ret) return ret; - return drm_get_pci_dev(pdev, ent, &kms_driver); + dev = drm_dev_alloc(&kms_driver, &pdev->dev); + if (IS_ERR(dev)) + return PTR_ERR(dev); + + ret = pci_enable_device(pdev); + if (ret) + goto err_free; + + dev->pdev = pdev; +#ifdef __alpha__ + dev->hose = pdev->sysdata; +#endif + + pci_set_drvdata(pdev, dev); + + if (pci_find_capability(dev->pdev, PCI_CAP_ID_AGP)) + dev->agp = drm_agp_init(dev); + if (dev->agp) { + dev->agp->agp_mtrr = arch_phys_wc_add( + dev->agp->agp_info.aper_base, + dev->agp->agp_info.aper_size * + 1024 * 1024); + } + + ret = drm_dev_register(dev, ent->driver_data); + if (ret) + goto err_agp; + + return 0; + +err_agp: + if (dev->agp) + arch_phys_wc_del(dev->agp->agp_mtrr); + kfree(dev->agp); + pci_disable_device(pdev); +err_free: + drm_dev_put(dev); + return ret; } static void @@ -575,7 +614,7 @@ radeon_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe, static struct drm_driver kms_driver = { .driver_features = - DRIVER_USE_AGP | DRIVER_GEM | DRIVER_RENDER, + DRIVER_GEM | DRIVER_RENDER, .load = radeon_driver_load_kms, .open = radeon_driver_open_kms, .postclose = radeon_driver_postclose_kms, diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index d24f23a81656..dd2f19b8022b 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -32,6 +32,7 @@ #include #include +#include #include #include #include @@ -77,6 +78,11 @@ void radeon_driver_unload_kms(struct drm_device *dev) radeon_modeset_fini(rdev); radeon_device_fini(rdev); + if (dev->agp) + arch_phys_wc_del(dev->agp->agp_mtrr); + kfree(dev->agp); + dev->agp = NULL; + done_free: kfree(rdev); dev->dev_private = NULL;