- Linux-stable-mirror - lists.linaro.org

Patch "drm/i915/perf: fix perf stream opening lock" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/i915/perf: fix perf stream opening lock to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-i915-perf-fix-perf-stream-opening-lock.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From f616f2830c1ed79245cfeca900f7e8a3b3c08c06 Mon Sep 17 00:00:00 2001 From: Lionel Landwerlin <lionel.g.landwerlin(a)intel.com> Date: Thu, 1 Mar 2018 11:06:13 +0000 Subject: drm/i915/perf: fix perf stream opening lock From: Lionel Landwerlin <lionel.g.landwerlin(a)intel.com> commit f616f2830c1ed79245cfeca900f7e8a3b3c08c06 upstream. We're seeing on CI that some contexts don't have the programmed OA period timer that directs the OA unit on how often to write reports. The issue is that we're not holding the drm lock from when we edit the context images down to when we set the exclusive_stream variable. This leaves a window for the deferred context allocation to call i915_oa_init_reg_state() that will not program the expected OA timer value, because we haven't set the exclusive_stream yet. v2: Drop need_lock from gen8_configure_all_contexts() (Matt) Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin(a)intel.com> Reviewed-by: Matthew Auld <matthew.auld(a)intel.com> Reviewed-by: Chris Wilson <chris(a)chris-wilson.co.uk> Fixes: 701f8231a2f ("drm/i915/perf: prune OA configs") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102254 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103715 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103755 Link: https://patchwork.freedesktop.org/patch/msgid/20180301110613.1737-1-lionel.… Cc: Jani Nikula <jani.nikula(a)linux.intel.com> Cc: Joonas Lahtinen <joonas.lahtinen(a)linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Cc: intel-gfx(a)lists.freedesktop.org Cc: <stable(a)vger.kernel.org> # v4.14+ (cherry picked from commit 41d3fdcd15d5ecf29cc73e8b79c2327ebb54b960) Signed-off-by: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/i915/i915_perf.c | 38 ++++++++++++-------------------------- 1 file changed, 12 insertions(+), 26 deletions(-) --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1300,9 +1300,8 @@ static void i915_oa_stream_destroy(struc */ mutex_lock(&dev_priv->drm.struct_mutex); dev_priv->perf.oa.exclusive_stream = NULL; - mutex_unlock(&dev_priv->drm.struct_mutex); - dev_priv->perf.oa.ops.disable_metric_set(dev_priv); + mutex_unlock(&dev_priv->drm.struct_mutex); free_oa_buffer(dev_priv); @@ -1754,22 +1753,13 @@ static int gen8_switch_to_updated_kernel * Note: it's only the RCS/Render context that has any OA state. */ static int gen8_configure_all_contexts(struct drm_i915_private *dev_priv, - const struct i915_oa_config *oa_config, - bool interruptible) + const struct i915_oa_config *oa_config) { struct i915_gem_context *ctx; int ret; unsigned int wait_flags = I915_WAIT_LOCKED; - if (interruptible) { - ret = i915_mutex_lock_interruptible(&dev_priv->drm); - if (ret) - return ret; - - wait_flags |= I915_WAIT_INTERRUPTIBLE; - } else { - mutex_lock(&dev_priv->drm.struct_mutex); - } + lockdep_assert_held(&dev_priv->drm.struct_mutex); /* Switch away from any user context. */ ret = gen8_switch_to_updated_kernel_context(dev_priv, oa_config); @@ -1817,8 +1807,6 @@ static int gen8_configure_all_contexts(s } out: - mutex_unlock(&dev_priv->drm.struct_mutex); - return ret; } @@ -1862,7 +1850,7 @@ static int gen8_enable_metric_set(struct * to make sure all slices/subslices are ON before writing to NOA * registers. */ - ret = gen8_configure_all_contexts(dev_priv, oa_config, true); + ret = gen8_configure_all_contexts(dev_priv, oa_config); if (ret) return ret; @@ -1877,7 +1865,7 @@ static int gen8_enable_metric_set(struct static void gen8_disable_metric_set(struct drm_i915_private *dev_priv) { /* Reset all contexts' slices/subslices configurations. */ - gen8_configure_all_contexts(dev_priv, NULL, false); + gen8_configure_all_contexts(dev_priv, NULL); I915_WRITE(GDT_CHICKEN_BITS, (I915_READ(GDT_CHICKEN_BITS) & ~GT_NOA_ENABLE)); @@ -2127,6 +2115,10 @@ static int i915_oa_stream_init(struct i9 if (ret) goto err_oa_buf_alloc; + ret = i915_mutex_lock_interruptible(&dev_priv->drm); + if (ret) + goto err_lock; + ret = dev_priv->perf.oa.ops.enable_metric_set(dev_priv, stream->oa_config); if (ret) @@ -2134,23 +2126,17 @@ static int i915_oa_stream_init(struct i9 stream->ops = &i915_oa_stream_ops; - /* Lock device for exclusive_stream access late because - * enable_metric_set() might lock as well on gen8+. - */ - ret = i915_mutex_lock_interruptible(&dev_priv->drm); - if (ret) - goto err_lock; - dev_priv->perf.oa.exclusive_stream = stream; mutex_unlock(&dev_priv->drm.struct_mutex); return 0; -err_lock: +err_enable: dev_priv->perf.oa.ops.disable_metric_set(dev_priv); + mutex_unlock(&dev_priv->drm.struct_mutex); -err_enable: +err_lock: free_oa_buffer(dev_priv); err_oa_buf_alloc: Patches currently in stable-queue which might be from lionel.g.landwerlin(a)intel.com are queue-4.14/drm-i915-perf-fix-perf-stream-opening-lock.patch

7 years, 4 months

1
0
0 0

Patch "drm/i915: Fix rsvd2 mask when out-fence is returned" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/i915: Fix rsvd2 mask when out-fence is returned to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-i915-fix-rsvd2-mask-when-out-fence-is-returned.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From b1b13780ab06ef8c770dd9cbe31dac549a11630e Mon Sep 17 00:00:00 2001 From: Daniele Ceraolo Spurio <daniele.ceraolospurio(a)intel.com> Date: Wed, 14 Feb 2018 11:18:25 -0800 Subject: drm/i915: Fix rsvd2 mask when out-fence is returned From: Daniele Ceraolo Spurio <daniele.ceraolospurio(a)intel.com> commit b1b13780ab06ef8c770dd9cbe31dac549a11630e upstream. GENMASK_ULL wants the high bit of the mask first. The current value cancels the in-fence when an out-fence is returned. Fixes: fec0445caa273 ("drm/i915: Support explicit fencing for execbuf") Testcase: igt/gem_exec_fence/keep-in-fence* Cc: Chris Wilson <chris(a)chris-wilson.co.uk> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio(a)intel.com> Reviewed-by: Chris Wilson <chris(a)chris-wilson.co.uk> Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Link: https://patchwork.freedesktop.org/patch/msgid/20180214191827.8465-1-daniele… Cc: <stable(a)vger.kernel.org> # v4.12+ (cherry picked from commit b6a88e4a804cf5a71159906e16df2c1fc7196f92) Signed-off-by: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/i915/i915_gem_execbuffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -2408,7 +2408,7 @@ err_request: if (out_fence) { if (err == 0) { fd_install(out_fence_fd, out_fence->file); - args->rsvd2 &= GENMASK_ULL(0, 31); /* keep in-fence */ + args->rsvd2 &= GENMASK_ULL(31, 0); /* keep in-fence */ args->rsvd2 |= (u64)out_fence_fd << 32; out_fence_fd = -1; } else { Patches currently in stable-queue which might be from daniele.ceraolospurio(a)intel.com are queue-4.14/drm-i915-fix-rsvd2-mask-when-out-fence-is-returned.patch

7 years, 4 months

1
0
0 0

Patch "drm/i915: Disable DC states around GMBUS on GLK" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/i915: Disable DC states around GMBUS on GLK to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-i915-disable-dc-states-around-gmbus-on-glk.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 156961ae7bdf6feb72778e8da83d321b273343fd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= <ville.syrjala(a)linux.intel.com> Date: Fri, 8 Dec 2017 23:37:36 +0200 Subject: drm/i915: Disable DC states around GMBUS on GLK MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Ville Syrjälä <ville.syrjala(a)linux.intel.com> commit 156961ae7bdf6feb72778e8da83d321b273343fd upstream. Prevent the DMC from destroying GMBUS transfers on GLK. GMBUS lives in PG1 so DC off is all we need. Cc: stable(a)vger.kernel.org Signed-off-by: Ville Syrjälä <ville.syrjala(a)linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20171208213739.16388-1-ville.… Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/i915/intel_runtime_pm.c | 1 + 1 file changed, 1 insertion(+) --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -1844,6 +1844,7 @@ void intel_display_power_put(struct drm_ CNL_DISPLAY_POWERWELL_2_POWER_DOMAINS | \ BIT_ULL(POWER_DOMAIN_MODESET) | \ BIT_ULL(POWER_DOMAIN_AUX_A) | \ + BIT_ULL(POWER_DOMAIN_GMBUS) | \ BIT_ULL(POWER_DOMAIN_INIT)) static const struct i915_power_well_ops i9xx_always_on_power_well_ops = { Patches currently in stable-queue which might be from ville.syrjala(a)linux.intel.com are queue-4.14/drm-i915-update-watermark-state-correctly-in-sanitize_watermarks.patch queue-4.14/drm-i915-disable-dc-states-around-gmbus-on-glk.patch

7 years, 4 months

1
0
0 0

Patch "drm/i915: Clear the in-use marker on execbuf failure" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/i915: Clear the in-use marker on execbuf failure to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-i915-clear-the-in-use-marker-on-execbuf-failure.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From e659d14ed48096f87a678e7ebbdf286a817b4d0e Mon Sep 17 00:00:00 2001 From: Chris Wilson <chris(a)chris-wilson.co.uk> Date: Mon, 19 Feb 2018 14:01:44 +0000 Subject: drm/i915: Clear the in-use marker on execbuf failure From: Chris Wilson <chris(a)chris-wilson.co.uk> commit e659d14ed48096f87a678e7ebbdf286a817b4d0e upstream. If we fail to unbind the vma (due to a signal on an active buffer that needs to be moved for the next execbuf), then we need to clear the persistent tracking state we setup for this execbuf. Fixes: c7c6e46f913b ("drm/i915: Convert execbuf to use struct-of-array packing for critical fields") Testcase: igt/gem_fenced_exec_thrash/no-spare-fences-busy* Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> Cc: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Cc: Joonas Lahtinen <joonas.lahtinen(a)linux.intel.com> Cc: <stable(a)vger.kernel.org> # v4.14+ Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin(a)intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180219140144.24004-1-chris@… (cherry picked from commit ed2f3532321083cf40e4da4e36234880e0136136) Signed-off-by: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/i915/i915_gem_execbuffer.c | 2 ++ 1 file changed, 2 insertions(+) --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -499,6 +499,8 @@ eb_add_vma(struct i915_execbuffer *eb, u list_add_tail(&vma->exec_link, &eb->unbound); if (drm_mm_node_allocated(&vma->node)) err = i915_vma_unbind(vma); + if (unlikely(err)) + vma->exec_flags = NULL; } return err; } Patches currently in stable-queue which might be from chris(a)chris-wilson.co.uk are queue-4.14/drm-i915-fix-rsvd2-mask-when-out-fence-is-returned.patch queue-4.14/drm-i915-perf-fix-perf-stream-opening-lock.patch queue-4.14/drm-i915-clear-the-in-use-marker-on-execbuf-failure.patch

7 years, 4 months

1
0
0 0

Patch "drm/i915: Always call to intel_display_set_init_power() in resume_early." has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/i915: Always call to intel_display_set_init_power() in resume_early. to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-i915-always-call-to-intel_display_set_init_power-in-resume_early.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From d13a8479f3584613b6aacbb793eae64578b8f69a Mon Sep 17 00:00:00 2001 From: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Date: Tue, 16 Jan 2018 16:53:24 +0100 Subject: drm/i915: Always call to intel_display_set_init_power() in resume_early. From: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> commit d13a8479f3584613b6aacbb793eae64578b8f69a upstream. intel_power_domains_init_hw() calls set_init_power, but when using runtime power management this call is skipped. This prevents hw readout from taking place. Signed-off-by: Maarten Lankhorst <maarten.lankhorst(a)linux.intel.com> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104172 Link: https://patchwork.freedesktop.org/patch/msgid/20180116155324.75120-1-maarte… Fixes: bc87229f323e ("drm/i915/skl: enable PC9/10 power states during suspend-to-idle") Cc: Nivedita Swaminathan <nivedita.swaminathan(a)intel.com> Cc: Imre Deak <imre.deak(a)intel.com> Cc: Patrik Jakobsson <patrik.jakobsson(a)linux.intel.com> Cc: Jani Nikula <jani.nikula(a)linux.intel.com> Cc: Joonas Lahtinen <joonas.lahtinen(a)linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Cc: <stable(a)vger.kernel.org> # v4.5+ Reviewed-by: Imre Deak <imre.deak(a)intel.com> (cherry picked from commit ac25dfed15d470d7f23dd817e965b54aa3f94a1e) Signed-off-by: Rodrigo Vivi <rodrigo.vivi(a)intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/i915/i915_drv.c | 2 ++ 1 file changed, 2 insertions(+) --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -1806,6 +1806,8 @@ static int i915_drm_resume_early(struct if (IS_GEN9_LP(dev_priv) || !(dev_priv->suspended_to_idle && dev_priv->csr.dmc_payload)) intel_power_domains_init_hw(dev_priv, true); + else + intel_display_set_init_power(dev_priv, true); i915_gem_sanitize(dev_priv); Patches currently in stable-queue which might be from maarten.lankhorst(a)linux.intel.com are queue-4.14/drm-i915-update-watermark-state-correctly-in-sanitize_watermarks.patch queue-4.14/drm-i915-always-call-to-intel_display_set_init_power-in-resume_early.patch

7 years, 4 months

1
0
0 0

Patch "drm/amdgpu: used cached pcie gen info for SI (v2)" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu: used cached pcie gen info for SI (v2) to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-used-cached-pcie-gen-info-for-si-v2.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 05656e5e4917a08296300dc0530aed1539202c25 Mon Sep 17 00:00:00 2001 From: Alex Deucher <alexander.deucher(a)amd.com> Date: Mon, 26 Feb 2018 11:05:10 -0500 Subject: drm/amdgpu: used cached pcie gen info for SI (v2) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Alex Deucher <alexander.deucher(a)amd.com> commit 05656e5e4917a08296300dc0530aed1539202c25 upstream. Rather than querying it every time we need it. Also fixes a crash in VM pass through if there is no root bridge because the cached value fetch already checks this properly. v2: fix includes Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=105244 Acked-by: Christian König <christian.koenig(a)amd.com> Reviewed-by: Rex Zhu<rezhu(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/si.c | 22 +++++++-------- drivers/gpu/drm/amd/amdgpu/si_dpm.c | 50 +++++++++--------------------------- 2 files changed, 23 insertions(+), 49 deletions(-) --- a/drivers/gpu/drm/amd/amdgpu/si.c +++ b/drivers/gpu/drm/amd/amdgpu/si.c @@ -31,6 +31,7 @@ #include "amdgpu_uvd.h" #include "amdgpu_vce.h" #include "atom.h" +#include "amd_pcie.h" #include "amdgpu_powerplay.h" #include "sid.h" #include "si_ih.h" @@ -1461,8 +1462,8 @@ static void si_pcie_gen3_enable(struct a { struct pci_dev *root = adev->pdev->bus->self; int bridge_pos, gpu_pos; - u32 speed_cntl, mask, current_data_rate; - int ret, i; + u32 speed_cntl, current_data_rate; + int i; u16 tmp16; if (pci_is_root_bus(adev->pdev->bus)) @@ -1474,23 +1475,20 @@ static void si_pcie_gen3_enable(struct a if (adev->flags & AMD_IS_APU) return; - ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask); - if (ret != 0) - return; - - if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80))) + if (!(adev->pm.pcie_gen_mask & (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 | + CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3))) return; speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL); current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >> LC_CURRENT_DATA_RATE_SHIFT; - if (mask & DRM_PCIE_SPEED_80) { + if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) { if (current_data_rate == 2) { DRM_INFO("PCIE gen 3 link speeds already enabled\n"); return; } DRM_INFO("enabling PCIE gen 3 link speeds, disable with amdgpu.pcie_gen2=0\n"); - } else if (mask & DRM_PCIE_SPEED_50) { + } else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2) { if (current_data_rate == 1) { DRM_INFO("PCIE gen 2 link speeds already enabled\n"); return; @@ -1506,7 +1504,7 @@ static void si_pcie_gen3_enable(struct a if (!gpu_pos) return; - if (mask & DRM_PCIE_SPEED_80) { + if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) { if (current_data_rate != 2) { u16 bridge_cfg, gpu_cfg; u16 bridge_cfg2, gpu_cfg2; @@ -1589,9 +1587,9 @@ static void si_pcie_gen3_enable(struct a pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16); tmp16 &= ~0xf; - if (mask & DRM_PCIE_SPEED_80) + if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) tmp16 |= 3; - else if (mask & DRM_PCIE_SPEED_50) + else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2) tmp16 |= 2; else tmp16 |= 1; --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c @@ -26,6 +26,7 @@ #include "amdgpu_pm.h" #include "amdgpu_dpm.h" #include "amdgpu_atombios.h" +#include "amd_pcie.h" #include "sid.h" #include "r600_dpm.h" #include "si_dpm.h" @@ -3332,29 +3333,6 @@ static void btc_apply_voltage_delta_rule } } -static enum amdgpu_pcie_gen r600_get_pcie_gen_support(struct amdgpu_device *adev, - u32 sys_mask, - enum amdgpu_pcie_gen asic_gen, - enum amdgpu_pcie_gen default_gen) -{ - switch (asic_gen) { - case AMDGPU_PCIE_GEN1: - return AMDGPU_PCIE_GEN1; - case AMDGPU_PCIE_GEN2: - return AMDGPU_PCIE_GEN2; - case AMDGPU_PCIE_GEN3: - return AMDGPU_PCIE_GEN3; - default: - if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == AMDGPU_PCIE_GEN3)) - return AMDGPU_PCIE_GEN3; - else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == AMDGPU_PCIE_GEN2)) - return AMDGPU_PCIE_GEN2; - else - return AMDGPU_PCIE_GEN1; - } - return AMDGPU_PCIE_GEN1; -} - static void r600_calculate_u_and_p(u32 i, u32 r_c, u32 p_b, u32 *p, u32 *u) { @@ -5028,10 +5006,11 @@ static int si_populate_smc_acpi_state(st table->ACPIState.levels[0].vddc.index, &table->ACPIState.levels[0].std_vddc); } - table->ACPIState.levels[0].gen2PCIE = (u8)r600_get_pcie_gen_support(adev, - si_pi->sys_pcie_mask, - si_pi->boot_pcie_gen, - AMDGPU_PCIE_GEN1); + table->ACPIState.levels[0].gen2PCIE = + (u8)amdgpu_get_pcie_gen_support(adev, + si_pi->sys_pcie_mask, + si_pi->boot_pcie_gen, + AMDGPU_PCIE_GEN1); if (si_pi->vddc_phase_shed_control) si_populate_phase_shedding_value(adev, @@ -7162,10 +7141,10 @@ static void si_parse_pplib_clock_info(st pl->vddc = le16_to_cpu(clock_info->si.usVDDC); pl->vddci = le16_to_cpu(clock_info->si.usVDDCI); pl->flags = le32_to_cpu(clock_info->si.ulFlags); - pl->pcie_gen = r600_get_pcie_gen_support(adev, - si_pi->sys_pcie_mask, - si_pi->boot_pcie_gen, - clock_info->si.ucPCIEGen); + pl->pcie_gen = amdgpu_get_pcie_gen_support(adev, + si_pi->sys_pcie_mask, + si_pi->boot_pcie_gen, + clock_info->si.ucPCIEGen); /* patch up vddc if necessary */ ret = si_get_leakage_voltage_from_leakage_index(adev, pl->vddc, @@ -7320,7 +7299,6 @@ static int si_dpm_init(struct amdgpu_dev struct si_power_info *si_pi; struct atom_clock_dividers dividers; int ret; - u32 mask; si_pi = kzalloc(sizeof(struct si_power_info), GFP_KERNEL); if (si_pi == NULL) @@ -7330,11 +7308,9 @@ static int si_dpm_init(struct amdgpu_dev eg_pi = &ni_pi->eg; pi = &eg_pi->rv7xx; - ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask); - if (ret) - si_pi->sys_pcie_mask = 0; - else - si_pi->sys_pcie_mask = mask; + si_pi->sys_pcie_mask = + (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK) >> + CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT; si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID; si_pi->boot_pcie_gen = si_get_current_pcie_speed(adev); Patches currently in stable-queue which might be from alexander.deucher(a)amd.com are queue-4.14/drm-amdgpu-notify-sbios-device-ready-before-send-request.patch queue-4.14/workqueue-allow-retrieval-of-current-task-s-work-struct.patch queue-4.14/drm-amd-powerplay-vega10-allow-mclk-switching-with-no-displays.patch queue-4.14/drm-radeon-fix-kv-harvesting.patch queue-4.14/drm-radeon-fix-deadlock-on-runtime-suspend.patch queue-4.14/drm-amdgpu-correct-max-uvd-handles.patch queue-4.14/drm-amd-powerplay-smu7-allow-mclk-switching-with-no-displays.patch queue-4.14/drm-amdgpu-fix-deadlock-on-runtime-suspend.patch queue-4.14/drm-amd-powerplay-fix-power-over-limit-on-fiji.patch queue-4.14/drm-allow-determining-if-current-task-is-output-poll-worker.patch queue-4.14/drm-amdgpu-used-cached-pcie-gen-info-for-si-v2.patch queue-4.14/revert-drm-radeon-pm-autoswitch-power-state-when-in-balanced-mode.patch queue-4.14/drm-amdgpu-fix-kv-harvesting.patch queue-4.14/drm-amdgpu-always-save-uvd-vcpu_bo-in-vm-mode.patch queue-4.14/drm-radeon-insist-on-32-bit-dma-for-cedar-on-ppc64-ppc64le.patch

7 years, 4 months

1
0
0 0

Patch "drm/amdgpu: Notify sbios device ready before send request" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu: Notify sbios device ready before send request to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-notify-sbios-device-ready-before-send-request.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 1bced75f4ab04bec55aecb57d99435dc6d0ae5a0 Mon Sep 17 00:00:00 2001 From: Rex Zhu <Rex.Zhu(a)amd.com> Date: Tue, 27 Feb 2018 18:20:53 +0800 Subject: drm/amdgpu: Notify sbios device ready before send request From: Rex Zhu <Rex.Zhu(a)amd.com> commit 1bced75f4ab04bec55aecb57d99435dc6d0ae5a0 upstream. it is required if a platform supports PCIe root complex core voltage reduction. After receiving this notification, SBIOS can apply default PCIe root complex power policy. Reviewed-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Rex Zhu <Rex.Zhu(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ 1 file changed, 3 insertions(+) --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -540,6 +540,9 @@ int amdgpu_acpi_pcie_performance_request size_t size; u32 retry = 3; + if (amdgpu_acpi_pcie_notify_device_ready(adev)) + return -EINVAL; + /* Get the device handle */ handle = ACPI_HANDLE(&adev->pdev->dev); if (!handle) Patches currently in stable-queue which might be from Rex.Zhu(a)amd.com are queue-4.14/drm-amdgpu-notify-sbios-device-ready-before-send-request.patch

7 years, 4 months

1
0
0 0

Patch "drm/amdgpu: fix KV harvesting" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu: fix KV harvesting to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-fix-kv-harvesting.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 545b0bcde7fbd3ee408fa842ea0731451dc4bd0a Mon Sep 17 00:00:00 2001 From: Alex Deucher <alexander.deucher(a)amd.com> Date: Thu, 1 Mar 2018 11:05:31 -0500 Subject: drm/amdgpu: fix KV harvesting MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Alex Deucher <alexander.deucher(a)amd.com> commit 545b0bcde7fbd3ee408fa842ea0731451dc4bd0a upstream. Always set the graphics values to the max for the asic type. E.g., some 1 RB chips are actually 1 RB chips, others are actually harvested 2 RB chips. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=99353 Reviewed-by: Christian König <christian.koenig(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c | 30 ++---------------------------- 1 file changed, 2 insertions(+), 28 deletions(-) --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -4387,34 +4387,8 @@ static void gfx_v7_0_gpu_early_init(stru case CHIP_KAVERI: adev->gfx.config.max_shader_engines = 1; adev->gfx.config.max_tile_pipes = 4; - if ((adev->pdev->device == 0x1304) || - (adev->pdev->device == 0x1305) || - (adev->pdev->device == 0x130C) || - (adev->pdev->device == 0x130F) || - (adev->pdev->device == 0x1310) || - (adev->pdev->device == 0x1311) || - (adev->pdev->device == 0x131C)) { - adev->gfx.config.max_cu_per_sh = 8; - adev->gfx.config.max_backends_per_se = 2; - } else if ((adev->pdev->device == 0x1309) || - (adev->pdev->device == 0x130A) || - (adev->pdev->device == 0x130D) || - (adev->pdev->device == 0x1313) || - (adev->pdev->device == 0x131D)) { - adev->gfx.config.max_cu_per_sh = 6; - adev->gfx.config.max_backends_per_se = 2; - } else if ((adev->pdev->device == 0x1306) || - (adev->pdev->device == 0x1307) || - (adev->pdev->device == 0x130B) || - (adev->pdev->device == 0x130E) || - (adev->pdev->device == 0x1315) || - (adev->pdev->device == 0x131B)) { - adev->gfx.config.max_cu_per_sh = 4; - adev->gfx.config.max_backends_per_se = 1; - } else { - adev->gfx.config.max_cu_per_sh = 3; - adev->gfx.config.max_backends_per_se = 1; - } + adev->gfx.config.max_cu_per_sh = 8; + adev->gfx.config.max_backends_per_se = 2; adev->gfx.config.max_sh_per_se = 1; adev->gfx.config.max_texture_channel_caches = 4; adev->gfx.config.max_gprs = 256; Patches currently in stable-queue which might be from alexander.deucher(a)amd.com are queue-4.14/drm-amdgpu-notify-sbios-device-ready-before-send-request.patch queue-4.14/workqueue-allow-retrieval-of-current-task-s-work-struct.patch queue-4.14/drm-amd-powerplay-vega10-allow-mclk-switching-with-no-displays.patch queue-4.14/drm-radeon-fix-kv-harvesting.patch queue-4.14/drm-radeon-fix-deadlock-on-runtime-suspend.patch queue-4.14/drm-amdgpu-correct-max-uvd-handles.patch queue-4.14/drm-amd-powerplay-smu7-allow-mclk-switching-with-no-displays.patch queue-4.14/drm-amdgpu-fix-deadlock-on-runtime-suspend.patch queue-4.14/drm-amd-powerplay-fix-power-over-limit-on-fiji.patch queue-4.14/drm-allow-determining-if-current-task-is-output-poll-worker.patch queue-4.14/drm-amdgpu-used-cached-pcie-gen-info-for-si-v2.patch queue-4.14/revert-drm-radeon-pm-autoswitch-power-state-when-in-balanced-mode.patch queue-4.14/drm-amdgpu-fix-kv-harvesting.patch queue-4.14/drm-amdgpu-always-save-uvd-vcpu_bo-in-vm-mode.patch queue-4.14/drm-radeon-insist-on-32-bit-dma-for-cedar-on-ppc64-ppc64le.patch

7 years, 4 months

1
0
0 0

Patch "drm/amdgpu: Fix deadlock on runtime suspend" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu: Fix deadlock on runtime suspend to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-fix-deadlock-on-runtime-suspend.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From aa0aad57909eb321746325951d66af88a83bc956 Mon Sep 17 00:00:00 2001 From: Lukas Wunner <lukas(a)wunner.de> Date: Sun, 11 Feb 2018 10:38:28 +0100 Subject: drm/amdgpu: Fix deadlock on runtime suspend From: Lukas Wunner <lukas(a)wunner.de> commit aa0aad57909eb321746325951d66af88a83bc956 upstream. amdgpu's ->runtime_suspend hook calls drm_kms_helper_poll_disable(), which waits for the output poll worker to finish if it's running. The output poll worker meanwhile calls pm_runtime_get_sync() in amdgpu's ->detect hooks, which waits for the ongoing suspend to finish, causing a deadlock. Fix by not acquiring a runtime PM ref if the ->detect hooks are called in the output poll worker's context. This is safe because the poll worker is only enabled while runtime active and we know that ->runtime_suspend waits for it to finish. Fixes: d38ceaf99ed0 ("drm/amdgpu: add core driver (v4)") Cc: stable(a)vger.kernel.org # v4.2+: 27d4ee03078a: workqueue: Allow retrieval of current task's work struct Cc: stable(a)vger.kernel.org # v4.2+: 25c058ccaf2e: drm: Allow determining if current task is output poll worker Cc: Alex Deucher <alexander.deucher(a)amd.com> Tested-by: Mike Lothian <mike(a)fireburn.co.uk> Reviewed-by: Lyude Paul <lyude(a)redhat.com> Signed-off-by: Lukas Wunner <lukas(a)wunner.de> Link: https://patchwork.freedesktop.org/patch/msgid/4c9bf72aacae1eef062bd134cd112… Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c | 58 ++++++++++++++++--------- 1 file changed, 38 insertions(+), 20 deletions(-) --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c @@ -739,9 +739,11 @@ amdgpu_connector_lvds_detect(struct drm_ enum drm_connector_status ret = connector_status_disconnected; int r; - r = pm_runtime_get_sync(connector->dev->dev); - if (r < 0) - return connector_status_disconnected; + if (!drm_kms_helper_is_poll_worker()) { + r = pm_runtime_get_sync(connector->dev->dev); + if (r < 0) + return connector_status_disconnected; + } if (encoder) { struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); @@ -760,8 +762,12 @@ amdgpu_connector_lvds_detect(struct drm_ /* check acpi lid status ??? */ amdgpu_connector_update_scratch_regs(connector, ret); - pm_runtime_mark_last_busy(connector->dev->dev); - pm_runtime_put_autosuspend(connector->dev->dev); + + if (!drm_kms_helper_is_poll_worker()) { + pm_runtime_mark_last_busy(connector->dev->dev); + pm_runtime_put_autosuspend(connector->dev->dev); + } + return ret; } @@ -871,9 +877,11 @@ amdgpu_connector_vga_detect(struct drm_c enum drm_connector_status ret = connector_status_disconnected; int r; - r = pm_runtime_get_sync(connector->dev->dev); - if (r < 0) - return connector_status_disconnected; + if (!drm_kms_helper_is_poll_worker()) { + r = pm_runtime_get_sync(connector->dev->dev); + if (r < 0) + return connector_status_disconnected; + } encoder = amdgpu_connector_best_single_encoder(connector); if (!encoder) @@ -927,8 +935,10 @@ amdgpu_connector_vga_detect(struct drm_c amdgpu_connector_update_scratch_regs(connector, ret); out: - pm_runtime_mark_last_busy(connector->dev->dev); - pm_runtime_put_autosuspend(connector->dev->dev); + if (!drm_kms_helper_is_poll_worker()) { + pm_runtime_mark_last_busy(connector->dev->dev); + pm_runtime_put_autosuspend(connector->dev->dev); + } return ret; } @@ -991,9 +1001,11 @@ amdgpu_connector_dvi_detect(struct drm_c enum drm_connector_status ret = connector_status_disconnected; bool dret = false, broken_edid = false; - r = pm_runtime_get_sync(connector->dev->dev); - if (r < 0) - return connector_status_disconnected; + if (!drm_kms_helper_is_poll_worker()) { + r = pm_runtime_get_sync(connector->dev->dev); + if (r < 0) + return connector_status_disconnected; + } if (!force && amdgpu_connector_check_hpd_status_unchanged(connector)) { ret = connector->status; @@ -1118,8 +1130,10 @@ out: amdgpu_connector_update_scratch_regs(connector, ret); exit: - pm_runtime_mark_last_busy(connector->dev->dev); - pm_runtime_put_autosuspend(connector->dev->dev); + if (!drm_kms_helper_is_poll_worker()) { + pm_runtime_mark_last_busy(connector->dev->dev); + pm_runtime_put_autosuspend(connector->dev->dev); + } return ret; } @@ -1362,9 +1376,11 @@ amdgpu_connector_dp_detect(struct drm_co struct drm_encoder *encoder = amdgpu_connector_best_single_encoder(connector); int r; - r = pm_runtime_get_sync(connector->dev->dev); - if (r < 0) - return connector_status_disconnected; + if (!drm_kms_helper_is_poll_worker()) { + r = pm_runtime_get_sync(connector->dev->dev); + if (r < 0) + return connector_status_disconnected; + } if (!force && amdgpu_connector_check_hpd_status_unchanged(connector)) { ret = connector->status; @@ -1432,8 +1448,10 @@ amdgpu_connector_dp_detect(struct drm_co amdgpu_connector_update_scratch_regs(connector, ret); out: - pm_runtime_mark_last_busy(connector->dev->dev); - pm_runtime_put_autosuspend(connector->dev->dev); + if (!drm_kms_helper_is_poll_worker()) { + pm_runtime_mark_last_busy(connector->dev->dev); + pm_runtime_put_autosuspend(connector->dev->dev); + } return ret; } Patches currently in stable-queue which might be from lukas(a)wunner.de are queue-4.14/workqueue-allow-retrieval-of-current-task-s-work-struct.patch queue-4.14/drm-radeon-fix-deadlock-on-runtime-suspend.patch queue-4.14/drm-amdgpu-fix-deadlock-on-runtime-suspend.patch queue-4.14/drm-allow-determining-if-current-task-is-output-poll-worker.patch queue-4.14/drm-nouveau-fix-deadlock-on-runtime-suspend.patch

7 years, 4 months

1
0
0 0

Patch "drm/amdgpu:Correct max uvd handles" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled drm/amdgpu:Correct max uvd handles to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: drm-amdgpu-correct-max-uvd-handles.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From 0e5ee33d2a54e4c55fe92857f23e1cbb0440d6de Mon Sep 17 00:00:00 2001 From: James Zhu <James.Zhu(a)amd.com> Date: Tue, 6 Mar 2018 14:43:50 -0500 Subject: drm/amdgpu:Correct max uvd handles MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: James Zhu <James.Zhu(a)amd.com> commit 0e5ee33d2a54e4c55fe92857f23e1cbb0440d6de upstream. Max uvd handles should use adev->uvd.max_handles instead of AMDGPU_MAX_UVD_HANDLES here. Signed-off-by: James Zhu <James.Zhu(a)amd.com> Reviewed-by: Leo Liu <leo.liu(a)amd.com> Reviewed-by: Christian König <christian.koenig(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -297,7 +297,7 @@ int amdgpu_uvd_suspend(struct amdgpu_dev if (atomic_read(&adev->uvd.handles[i])) break; - if (i == AMDGPU_MAX_UVD_HANDLES) + if (i == adev->uvd.max_handles) return 0; cancel_delayed_work_sync(&adev->uvd.idle_work); Patches currently in stable-queue which might be from James.Zhu(a)amd.com are queue-4.14/drm-amdgpu-correct-max-uvd-handles.patch queue-4.14/drm-amdgpu-always-save-uvd-vcpu_bo-in-vm-mode.patch

7 years, 4 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror