| Message ID | 20230329095933.1203559-1-kai.heng.feng@canonical.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [1/2] drm/amdgpu: Reset GPU on S0ix when device supports BOCO | expand |
On Wed, Mar 29, 2023 at 6:00 AM Kai-Heng Feng <kai.heng.feng@canonical.com> wrote: > > When the power is lost due to ACPI power resources being turned off, the > driver should reset the GPU so it can work anew. > > First, _PR3 support of the hierarchy needs to be found correctly. Since > the GPU on some discrete GFX cards is behind a PCIe switch, checking the > _PR3 on downstream port alone is not enough, as the _PR3 can associate > to the root port above the PCIe switch. > > Once the _PR3 is found and BOCO support is correctly marked, use that > information to inform the GPU should be reset. This solves an issue that > system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold > is supported for the GFX slot. I don't think we need to reset the GPU. If the power is turned off, a reset shouldn't be necessary. The reset is only necessary when the power is not turned off to put the GPU into a known good state. It should be in that state already if the power is turn off. It sounds like the device is not actually getting powered off. Alex > > Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- > 3 files changed, 14 insertions(+), 8 deletions(-) > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > index 60b1857f469e..407456ac0e84 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) > if (amdgpu_sriov_vf(adev)) > return false; > > + if (amdgpu_device_supports_boco(adev_to_drm(adev))) > + return true; > + > #if IS_ENABLED(CONFIG_SUSPEND) > return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; > #else > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > index f5658359ff5c..d56b7a2bafa6 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) > > if (!(adev->flags & AMD_IS_APU)) { > parent = pci_upstream_bridge(adev->pdev); > - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; > + do { > + if (pci_pr3_present(parent)) { > + adev->has_pr3 = true; > + break; > + } > + } while ((parent = pci_upstream_bridge(parent))); > } > > amdgpu_amdkfd_device_probe(adev); > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > index ba5def374368..5d81fcac4b0a 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) > struct drm_device *drm_dev = dev_get_drvdata(dev); > struct amdgpu_device *adev = drm_to_adev(drm_dev); > > - if (amdgpu_acpi_is_s0ix_active(adev)) > - adev->in_s0ix = true; > - else if (amdgpu_acpi_is_s3_active(adev)) > + if (amdgpu_acpi_is_s3_active(adev) || > + amdgpu_device_supports_boco(drm_dev)) > adev->in_s3 = true; > + else if (amdgpu_acpi_is_s0ix_active(adev)) > + adev->in_s0ix = true; > if (!adev->in_s0ix && !adev->in_s3) > return 0; > return amdgpu_device_suspend(drm_dev, true); > @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) > adev->no_hw_access = true; > > r = amdgpu_device_resume(drm_dev, true); > - if (amdgpu_acpi_is_s0ix_active(adev)) > - adev->in_s0ix = false; > - else > - adev->in_s3 = false; > + adev->in_s0ix = adev->in_s3 = false; > return r; > } > > -- > 2.34.1 >
On 3/29/23 04:59, Kai-Heng Feng wrote: > When the power is lost due to ACPI power resources being turned off, the > driver should reset the GPU so it can work anew. > > First, _PR3 support of the hierarchy needs to be found correctly. Since > the GPU on some discrete GFX cards is behind a PCIe switch, checking the > _PR3 on downstream port alone is not enough, as the _PR3 can associate > to the root port above the PCIe switch. I think this should be split into two commits: * One of them to look at _PR3 further up in hierarchy to fix indication for BOCO support. * One to adjust policy for whether to reset > Once the _PR3 is found and BOCO support is correctly marked, use that > information to inform the GPU should be reset. This solves an issue that > system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold > is supported for the GFX slot. I'm worried this is still papering over an underlying issue with L0s handling on ALD + Navi1x/Navi2x. Also, what about runtime suspend? If you unplug the monitor from this dGPU and interact with it over SSH it should go into runtime suspend. Is it working properly for that case now? > > Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- > 3 files changed, 14 insertions(+), 8 deletions(-) > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > index 60b1857f469e..407456ac0e84 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) > if (amdgpu_sriov_vf(adev)) > return false; > > + if (amdgpu_device_supports_boco(adev_to_drm(adev))) > + return true; > + > #if IS_ENABLED(CONFIG_SUSPEND) > return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; > #else > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > index f5658359ff5c..d56b7a2bafa6 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) > > if (!(adev->flags & AMD_IS_APU)) { > parent = pci_upstream_bridge(adev->pdev); > - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; > + do { > + if (pci_pr3_present(parent)) { > + adev->has_pr3 = true; > + break; > + } > + } while ((parent = pci_upstream_bridge(parent))); > } > > amdgpu_amdkfd_device_probe(adev); > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > index ba5def374368..5d81fcac4b0a 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) > struct drm_device *drm_dev = dev_get_drvdata(dev); > struct amdgpu_device *adev = drm_to_adev(drm_dev); > > - if (amdgpu_acpi_is_s0ix_active(adev)) > - adev->in_s0ix = true; > - else if (amdgpu_acpi_is_s3_active(adev)) > + if (amdgpu_acpi_is_s3_active(adev) || > + amdgpu_device_supports_boco(drm_dev)) > adev->in_s3 = true; > + else if (amdgpu_acpi_is_s0ix_active(adev)) > + adev->in_s0ix = true; > if (!adev->in_s0ix && !adev->in_s3) > return 0; > return amdgpu_device_suspend(drm_dev, true); > @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) > adev->no_hw_access = true; > > r = amdgpu_device_resume(drm_dev, true); > - if (amdgpu_acpi_is_s0ix_active(adev)) > - adev->in_s0ix = false; > - else > - adev->in_s3 = false; > + adev->in_s0ix = adev->in_s3 = false; > return r; > } >
On Wed, Mar 29, 2023 at 9:21 PM Alex Deucher <alexdeucher@gmail.com> wrote: > > On Wed, Mar 29, 2023 at 6:00 AM Kai-Heng Feng > <kai.heng.feng@canonical.com> wrote: > > > > When the power is lost due to ACPI power resources being turned off, the > > driver should reset the GPU so it can work anew. > > > > First, _PR3 support of the hierarchy needs to be found correctly. Since > > the GPU on some discrete GFX cards is behind a PCIe switch, checking the > > _PR3 on downstream port alone is not enough, as the _PR3 can associate > > to the root port above the PCIe switch. > > > > Once the _PR3 is found and BOCO support is correctly marked, use that > > information to inform the GPU should be reset. This solves an issue that > > system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold > > is supported for the GFX slot. > > I don't think we need to reset the GPU. If the power is turned off, a > reset shouldn't be necessary. The reset is only necessary when the > power is not turned off to put the GPU into a known good state. It > should be in that state already if the power is turn off. It sounds > like the device is not actually getting powered off. I had the impression that the GPU gets reset because S3 turned the power rail off. So the actual intention for GPU reset is because S3 doesn't guarantee the power is being turned off? Kai-Heng > > Alex > > > > > Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 > > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > --- > > drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ > > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- > > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- > > 3 files changed, 14 insertions(+), 8 deletions(-) > > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > index 60b1857f469e..407456ac0e84 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) > > if (amdgpu_sriov_vf(adev)) > > return false; > > > > + if (amdgpu_device_supports_boco(adev_to_drm(adev))) > > + return true; > > + > > #if IS_ENABLED(CONFIG_SUSPEND) > > return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; > > #else > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > index f5658359ff5c..d56b7a2bafa6 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) > > > > if (!(adev->flags & AMD_IS_APU)) { > > parent = pci_upstream_bridge(adev->pdev); > > - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; > > + do { > > + if (pci_pr3_present(parent)) { > > + adev->has_pr3 = true; > > + break; > > + } > > + } while ((parent = pci_upstream_bridge(parent))); > > } > > > > amdgpu_amdkfd_device_probe(adev); > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > index ba5def374368..5d81fcac4b0a 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) > > struct drm_device *drm_dev = dev_get_drvdata(dev); > > struct amdgpu_device *adev = drm_to_adev(drm_dev); > > > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > - adev->in_s0ix = true; > > - else if (amdgpu_acpi_is_s3_active(adev)) > > + if (amdgpu_acpi_is_s3_active(adev) || > > + amdgpu_device_supports_boco(drm_dev)) > > adev->in_s3 = true; > > + else if (amdgpu_acpi_is_s0ix_active(adev)) > > + adev->in_s0ix = true; > > if (!adev->in_s0ix && !adev->in_s3) > > return 0; > > return amdgpu_device_suspend(drm_dev, true); > > @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) > > adev->no_hw_access = true; > > > > r = amdgpu_device_resume(drm_dev, true); > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > - adev->in_s0ix = false; > > - else > > - adev->in_s3 = false; > > + adev->in_s0ix = adev->in_s3 = false; > > return r; > > } > > > > -- > > 2.34.1 > >
On Wed, Mar 29, 2023 at 8:49 PM Kai-Heng Feng <kai.heng.feng@canonical.com> wrote: > > On Wed, Mar 29, 2023 at 9:21 PM Alex Deucher <alexdeucher@gmail.com> wrote: > > > > On Wed, Mar 29, 2023 at 6:00 AM Kai-Heng Feng > > <kai.heng.feng@canonical.com> wrote: > > > > > > When the power is lost due to ACPI power resources being turned off, the > > > driver should reset the GPU so it can work anew. > > > > > > First, _PR3 support of the hierarchy needs to be found correctly. Since > > > the GPU on some discrete GFX cards is behind a PCIe switch, checking the > > > _PR3 on downstream port alone is not enough, as the _PR3 can associate > > > to the root port above the PCIe switch. > > > > > > Once the _PR3 is found and BOCO support is correctly marked, use that > > > information to inform the GPU should be reset. This solves an issue that > > > system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold > > > is supported for the GFX slot. > > > > I don't think we need to reset the GPU. If the power is turned off, a > > reset shouldn't be necessary. The reset is only necessary when the > > power is not turned off to put the GPU into a known good state. It > > should be in that state already if the power is turn off. It sounds > > like the device is not actually getting powered off. > > I had the impression that the GPU gets reset because S3 turned the > power rail off. > > So the actual intention for GPU reset is because S3 doesn't guarantee > the power is being turned off? For S4, the reset in freeze is there because once the boot kernel transitions to the hibernated kernel, we need the reset to bring the GPU back to a known state. On dGPUs at least there are some engines that can only be initialized once and then require a reset to be initialized again. The one in suspend was originally there to deal with aborted suspends where we'd need to reset the GPU for the same reason as S4. However, it no longer really serves much purpose since it got moved to noirq and it could probably be dropped. Alex > > Kai-Heng > > > > > Alex > > > > > > > > Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") > > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 > > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 > > > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > > --- > > > drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ > > > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- > > > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- > > > 3 files changed, 14 insertions(+), 8 deletions(-) > > > > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > > index 60b1857f469e..407456ac0e84 100644 > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > > @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) > > > if (amdgpu_sriov_vf(adev)) > > > return false; > > > > > > + if (amdgpu_device_supports_boco(adev_to_drm(adev))) > > > + return true; > > > + > > > #if IS_ENABLED(CONFIG_SUSPEND) > > > return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; > > > #else > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > > index f5658359ff5c..d56b7a2bafa6 100644 > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > > @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) > > > > > > if (!(adev->flags & AMD_IS_APU)) { > > > parent = pci_upstream_bridge(adev->pdev); > > > - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; > > > + do { > > > + if (pci_pr3_present(parent)) { > > > + adev->has_pr3 = true; > > > + break; > > > + } > > > + } while ((parent = pci_upstream_bridge(parent))); > > > } > > > > > > amdgpu_amdkfd_device_probe(adev); > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > > index ba5def374368..5d81fcac4b0a 100644 > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > > @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) > > > struct drm_device *drm_dev = dev_get_drvdata(dev); > > > struct amdgpu_device *adev = drm_to_adev(drm_dev); > > > > > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > > - adev->in_s0ix = true; > > > - else if (amdgpu_acpi_is_s3_active(adev)) > > > + if (amdgpu_acpi_is_s3_active(adev) || > > > + amdgpu_device_supports_boco(drm_dev)) > > > adev->in_s3 = true; > > > + else if (amdgpu_acpi_is_s0ix_active(adev)) > > > + adev->in_s0ix = true; > > > if (!adev->in_s0ix && !adev->in_s3) > > > return 0; > > > return amdgpu_device_suspend(drm_dev, true); > > > @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) > > > adev->no_hw_access = true; > > > > > > r = amdgpu_device_resume(drm_dev, true); > > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > > - adev->in_s0ix = false; > > > - else > > > - adev->in_s3 = false; > > > + adev->in_s0ix = adev->in_s3 = false; > > > return r; > > > } > > > > > > -- > > > 2.34.1 > > >
On Wed, Mar 29, 2023 at 9:23 PM Mario Limonciello <mario.limonciello@amd.com> wrote: > > > On 3/29/23 04:59, Kai-Heng Feng wrote: > > When the power is lost due to ACPI power resources being turned off, the > > driver should reset the GPU so it can work anew. > > > > First, _PR3 support of the hierarchy needs to be found correctly. Since > > the GPU on some discrete GFX cards is behind a PCIe switch, checking the > > _PR3 on downstream port alone is not enough, as the _PR3 can associate > > to the root port above the PCIe switch. > > I think this should be split into two commits: > > * One of them to look at _PR3 further up in hierarchy to fix indication > for BOCO support. Yes, this part can be split up. > > * One to adjust policy for whether to reset IIUC, the GPU only needs to be reset when the power status isn't certain? Assuming power resources in _PR3 are really disabled, GPU is already reset by itself. That means reset shouldn't be necessary for D3cold, am I understanding it correctly? However, this is a desktop plugged with GFX card that has external power, does that assumption still stand? Perform resetting on D3cold can cover this scenario. > > > > Once the _PR3 is found and BOCO support is correctly marked, use that > > information to inform the GPU should be reset. This solves an issue that > > system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold > > is supported for the GFX slot. > > I'm worried this is still papering over an underlying issue with L0s > handling on ALD + Navi1x/Navi2x. Is it possible to get the ASIC's ASPM parameter under Windows? Knowing the difference can be useful. > > Also, what about runtime suspend? If you unplug the monitor from this > dGPU and interact with it over SSH it should go into runtime suspend. > > Is it working properly for that case now? Thanks for the tip. Runtime resume doesn't work at all: [ 1087.601631] pcieport 0000:00:01.0: power state changed by ACPI to D0 [ 1087.613820] pcieport 0000:00:01.0: restoring config space at offset 0x2c (was 0x43, writing 0x43) [ 1087.613835] pcieport 0000:00:01.0: restoring config space at offset 0x28 (was 0x41, writing 0x41) [ 1087.613841] pcieport 0000:00:01.0: restoring config space at offset 0x24 (was 0xfff10001, writing 0xfff10001) [ 1087.613978] pcieport 0000:00:01.0: PME# disabled [ 1087.613984] pcieport 0000:00:01.0: waiting 100 ms for downstream link, after activation [ 1089.330956] pcieport 0000:01:00.0: not ready 1023ms after resume; giving up [ 1089.373036] pcieport 0000:01:00.0: Unable to change power state from D3cold to D0, device inaccessible After a short while the whole system froze. So the upstream port of GFX's PCIe switch cannot be powered on again. Kai-Heng > > > > > Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 > > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > --- > > drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ > > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- > > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- > > 3 files changed, 14 insertions(+), 8 deletions(-) > > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > index 60b1857f469e..407456ac0e84 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) > > if (amdgpu_sriov_vf(adev)) > > return false; > > > > + if (amdgpu_device_supports_boco(adev_to_drm(adev))) > > + return true; > > + > > #if IS_ENABLED(CONFIG_SUSPEND) > > return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; > > #else > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > index f5658359ff5c..d56b7a2bafa6 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) > > > > if (!(adev->flags & AMD_IS_APU)) { > > parent = pci_upstream_bridge(adev->pdev); > > - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; > > + do { > > + if (pci_pr3_present(parent)) { > > + adev->has_pr3 = true; > > + break; > > + } > > + } while ((parent = pci_upstream_bridge(parent))); > > } > > > > amdgpu_amdkfd_device_probe(adev); > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > index ba5def374368..5d81fcac4b0a 100644 > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) > > struct drm_device *drm_dev = dev_get_drvdata(dev); > > struct amdgpu_device *adev = drm_to_adev(drm_dev); > > > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > - adev->in_s0ix = true; > > - else if (amdgpu_acpi_is_s3_active(adev)) > > + if (amdgpu_acpi_is_s3_active(adev) || > > + amdgpu_device_supports_boco(drm_dev)) > > adev->in_s3 = true; > > + else if (amdgpu_acpi_is_s0ix_active(adev)) > > + adev->in_s0ix = true; > > if (!adev->in_s0ix && !adev->in_s3) > > return 0; > > return amdgpu_device_suspend(drm_dev, true); > > @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) > > adev->no_hw_access = true; > > > > r = amdgpu_device_resume(drm_dev, true); > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > - adev->in_s0ix = false; > > - else > > - adev->in_s3 = false; > > + adev->in_s0ix = adev->in_s3 = false; > > return r; > > } > >
On 3/29/23 22:36, Kai-Heng Feng wrote: > On Wed, Mar 29, 2023 at 9:23 PM Mario Limonciello > <mario.limonciello@amd.com> wrote: >> >> On 3/29/23 04:59, Kai-Heng Feng wrote: >>> When the power is lost due to ACPI power resources being turned off, the >>> driver should reset the GPU so it can work anew. >>> >>> First, _PR3 support of the hierarchy needs to be found correctly. Since >>> the GPU on some discrete GFX cards is behind a PCIe switch, checking the >>> _PR3 on downstream port alone is not enough, as the _PR3 can associate >>> to the root port above the PCIe switch. >> I think this should be split into two commits: >> >> * One of them to look at _PR3 further up in hierarchy to fix indication >> for BOCO support. > Yes, this part can be split up. > >> * One to adjust policy for whether to reset > IIUC, the GPU only needs to be reset when the power status isn't certain? Yeah; think of reset as a particular case that all hardware isn't initialized. > Assuming power resources in _PR3 are really disabled, GPU is already > reset by itself. That means reset shouldn't be necessary for D3cold, > am I understanding it correctly? Can we see the acpidump for this system? > However, this is a desktop plugged with GFX card that has external > power, does that assumption still stand? Perform resetting on D3cold > can cover this scenario. Are you sure it's going to D3cold? Or is it D3hot? What does _S0W have in this design? >> >>> Once the _PR3 is found and BOCO support is correctly marked, use that >>> information to inform the GPU should be reset. This solves an issue that >>> system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold >>> is supported for the GFX slot. >> I'm worried this is still papering over an underlying issue with L0s >> handling on ALD + Navi1x/Navi2x. > Is it possible to get the ASIC's ASPM parameter under Windows? Knowing > the difference can be useful. Evan is in discussion with Windows guys about this issue too. > >> Also, what about runtime suspend? If you unplug the monitor from this >> dGPU and interact with it over SSH it should go into runtime suspend. >> >> Is it working properly for that case now? > Thanks for the tip. Runtime resume doesn't work at all: > [ 1087.601631] pcieport 0000:00:01.0: power state changed by ACPI to D0 > [ 1087.613820] pcieport 0000:00:01.0: restoring config space at offset > 0x2c (was 0x43, writing 0x43) > [ 1087.613835] pcieport 0000:00:01.0: restoring config space at offset > 0x28 (was 0x41, writing 0x41) > [ 1087.613841] pcieport 0000:00:01.0: restoring config space at offset > 0x24 (was 0xfff10001, writing 0xfff10001) > [ 1087.613978] pcieport 0000:00:01.0: PME# disabled > [ 1087.613984] pcieport 0000:00:01.0: waiting 100 ms for downstream > link, after activation > [ 1089.330956] pcieport 0000:01:00.0: not ready 1023ms after resume; giving up > [ 1089.373036] pcieport 0000:01:00.0: Unable to change power state > from D3cold to D0, device inaccessible > > After a short while the whole system froze. > > So the upstream port of GFX's PCIe switch cannot be powered on again. What is the state of the kernel patches while doing this test? Specifically does this happen without amdgpu.aspm set? Or this happens no matter what? > > Kai-Heng > >>> Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") >>> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 >>> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 >>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> >>> --- >>> drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ >>> drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- >>> drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- >>> 3 files changed, 14 insertions(+), 8 deletions(-) >>> >>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c >>> index 60b1857f469e..407456ac0e84 100644 >>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c >>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c >>> @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) >>> if (amdgpu_sriov_vf(adev)) >>> return false; >>> >>> + if (amdgpu_device_supports_boco(adev_to_drm(adev))) >>> + return true; >>> + >>> #if IS_ENABLED(CONFIG_SUSPEND) >>> return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; >>> #else >>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c >>> index f5658359ff5c..d56b7a2bafa6 100644 >>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c >>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c >>> @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) >>> >>> if (!(adev->flags & AMD_IS_APU)) { >>> parent = pci_upstream_bridge(adev->pdev); >>> - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; >>> + do { >>> + if (pci_pr3_present(parent)) { >>> + adev->has_pr3 = true; >>> + break; >>> + } >>> + } while ((parent = pci_upstream_bridge(parent))); >>> } >>> >>> amdgpu_amdkfd_device_probe(adev); >>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c >>> index ba5def374368..5d81fcac4b0a 100644 >>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c >>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c >>> @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) >>> struct drm_device *drm_dev = dev_get_drvdata(dev); >>> struct amdgpu_device *adev = drm_to_adev(drm_dev); >>> >>> - if (amdgpu_acpi_is_s0ix_active(adev)) >>> - adev->in_s0ix = true; >>> - else if (amdgpu_acpi_is_s3_active(adev)) >>> + if (amdgpu_acpi_is_s3_active(adev) || >>> + amdgpu_device_supports_boco(drm_dev)) >>> adev->in_s3 = true; >>> + else if (amdgpu_acpi_is_s0ix_active(adev)) >>> + adev->in_s0ix = true; >>> if (!adev->in_s0ix && !adev->in_s3) >>> return 0; >>> return amdgpu_device_suspend(drm_dev, true); >>> @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) >>> adev->no_hw_access = true; >>> >>> r = amdgpu_device_resume(drm_dev, true); >>> - if (amdgpu_acpi_is_s0ix_active(adev)) >>> - adev->in_s0ix = false; >>> - else >>> - adev->in_s3 = false; >>> + adev->in_s0ix = adev->in_s3 = false; >>> return r; >>> } >>>
On Wed, Mar 29, 2023 at 11:36 PM Kai-Heng Feng <kai.heng.feng@canonical.com> wrote: > > On Wed, Mar 29, 2023 at 9:23 PM Mario Limonciello > <mario.limonciello@amd.com> wrote: > > > > > > On 3/29/23 04:59, Kai-Heng Feng wrote: > > > When the power is lost due to ACPI power resources being turned off, the > > > driver should reset the GPU so it can work anew. > > > > > > First, _PR3 support of the hierarchy needs to be found correctly. Since > > > the GPU on some discrete GFX cards is behind a PCIe switch, checking the > > > _PR3 on downstream port alone is not enough, as the _PR3 can associate > > > to the root port above the PCIe switch. > > > > I think this should be split into two commits: > > > > * One of them to look at _PR3 further up in hierarchy to fix indication > > for BOCO support. > > Yes, this part can be split up. > > > > > * One to adjust policy for whether to reset > > IIUC, the GPU only needs to be reset when the power status isn't certain? > > Assuming power resources in _PR3 are really disabled, GPU is already > reset by itself. That means reset shouldn't be necessary for D3cold, > am I understanding it correctly? Right, if D3cold actually works, then no reset is necessary. > > However, this is a desktop plugged with GFX card that has external > power, does that assumption still stand? Perform resetting on D3cold > can cover this scenario. BOCO is generally only available on laptops and all-in-one systems where the dGPU is integrated into the platform. Power to the dGPU is controlled by a GPIO which is toggled by the ACPI _PR3 method for the device. There is an ATPX method on all platforms which support BOCO. Since this is an AIB in a desktop system, I doubt it actually supports D3Cold. For desktop systems, we have what we call BACO where the driver controls power to everything on the GPU except the bus interface. In the BACO case, we can turn off the GPU, but the device still shows up on the PCI bus. For BOCO, the device is completely powered down and disappears from the PCI bus. Alex > > > > > > > > Once the _PR3 is found and BOCO support is correctly marked, use that > > > information to inform the GPU should be reset. This solves an issue that > > > system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold > > > is supported for the GFX slot. > > > > I'm worried this is still papering over an underlying issue with L0s > > handling on ALD + Navi1x/Navi2x. > > Is it possible to get the ASIC's ASPM parameter under Windows? Knowing > the difference can be useful. > > > > > Also, what about runtime suspend? If you unplug the monitor from this > > dGPU and interact with it over SSH it should go into runtime suspend. > > > > Is it working properly for that case now? > > Thanks for the tip. Runtime resume doesn't work at all: > [ 1087.601631] pcieport 0000:00:01.0: power state changed by ACPI to D0 > [ 1087.613820] pcieport 0000:00:01.0: restoring config space at offset > 0x2c (was 0x43, writing 0x43) > [ 1087.613835] pcieport 0000:00:01.0: restoring config space at offset > 0x28 (was 0x41, writing 0x41) > [ 1087.613841] pcieport 0000:00:01.0: restoring config space at offset > 0x24 (was 0xfff10001, writing 0xfff10001) > [ 1087.613978] pcieport 0000:00:01.0: PME# disabled > [ 1087.613984] pcieport 0000:00:01.0: waiting 100 ms for downstream > link, after activation > [ 1089.330956] pcieport 0000:01:00.0: not ready 1023ms after resume; giving up > [ 1089.373036] pcieport 0000:01:00.0: Unable to change power state > from D3cold to D0, device inaccessible > > After a short while the whole system froze. > > So the upstream port of GFX's PCIe switch cannot be powered on again. > > Kai-Heng > > > > > > > > > Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") > > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 > > > Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 > > > Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > > --- > > > drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ > > > drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- > > > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- > > > 3 files changed, 14 insertions(+), 8 deletions(-) > > > > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > > index 60b1857f469e..407456ac0e84 100644 > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c > > > @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) > > > if (amdgpu_sriov_vf(adev)) > > > return false; > > > > > > + if (amdgpu_device_supports_boco(adev_to_drm(adev))) > > > + return true; > > > + > > > #if IS_ENABLED(CONFIG_SUSPEND) > > > return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; > > > #else > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > > index f5658359ff5c..d56b7a2bafa6 100644 > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c > > > @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) > > > > > > if (!(adev->flags & AMD_IS_APU)) { > > > parent = pci_upstream_bridge(adev->pdev); > > > - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; > > > + do { > > > + if (pci_pr3_present(parent)) { > > > + adev->has_pr3 = true; > > > + break; > > > + } > > > + } while ((parent = pci_upstream_bridge(parent))); > > > } > > > > > > amdgpu_amdkfd_device_probe(adev); > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > > index ba5def374368..5d81fcac4b0a 100644 > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > > > @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) > > > struct drm_device *drm_dev = dev_get_drvdata(dev); > > > struct amdgpu_device *adev = drm_to_adev(drm_dev); > > > > > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > > - adev->in_s0ix = true; > > > - else if (amdgpu_acpi_is_s3_active(adev)) > > > + if (amdgpu_acpi_is_s3_active(adev) || > > > + amdgpu_device_supports_boco(drm_dev)) > > > adev->in_s3 = true; > > > + else if (amdgpu_acpi_is_s0ix_active(adev)) > > > + adev->in_s0ix = true; > > > if (!adev->in_s0ix && !adev->in_s3) > > > return 0; > > > return amdgpu_device_suspend(drm_dev, true); > > > @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) > > > adev->no_hw_access = true; > > > > > > r = amdgpu_device_resume(drm_dev, true); > > > - if (amdgpu_acpi_is_s0ix_active(adev)) > > > - adev->in_s0ix = false; > > > - else > > > - adev->in_s3 = false; > > > + adev->in_s0ix = adev->in_s3 = false; > > > return r; > > > } > > >
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 60b1857f469e..407456ac0e84 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -987,6 +987,9 @@ bool amdgpu_acpi_should_gpu_reset(struct amdgpu_device *adev) if (amdgpu_sriov_vf(adev)) return false; + if (amdgpu_device_supports_boco(adev_to_drm(adev))) + return true; + #if IS_ENABLED(CONFIG_SUSPEND) return pm_suspend_target_state != PM_SUSPEND_TO_IDLE; #else diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index f5658359ff5c..d56b7a2bafa6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -2181,7 +2181,12 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) if (!(adev->flags & AMD_IS_APU)) { parent = pci_upstream_bridge(adev->pdev); - adev->has_pr3 = parent ? pci_pr3_present(parent) : false; + do { + if (pci_pr3_present(parent)) { + adev->has_pr3 = true; + break; + } + } while ((parent = pci_upstream_bridge(parent))); } amdgpu_amdkfd_device_probe(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index ba5def374368..5d81fcac4b0a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -2415,10 +2415,11 @@ static int amdgpu_pmops_suspend(struct device *dev) struct drm_device *drm_dev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(drm_dev); - if (amdgpu_acpi_is_s0ix_active(adev)) - adev->in_s0ix = true; - else if (amdgpu_acpi_is_s3_active(adev)) + if (amdgpu_acpi_is_s3_active(adev) || + amdgpu_device_supports_boco(drm_dev)) adev->in_s3 = true; + else if (amdgpu_acpi_is_s0ix_active(adev)) + adev->in_s0ix = true; if (!adev->in_s0ix && !adev->in_s3) return 0; return amdgpu_device_suspend(drm_dev, true); @@ -2449,10 +2450,7 @@ static int amdgpu_pmops_resume(struct device *dev) adev->no_hw_access = true; r = amdgpu_device_resume(drm_dev, true); - if (amdgpu_acpi_is_s0ix_active(adev)) - adev->in_s0ix = false; - else - adev->in_s3 = false; + adev->in_s0ix = adev->in_s3 = false; return r; }
When the power is lost due to ACPI power resources being turned off, the driver should reset the GPU so it can work anew. First, _PR3 support of the hierarchy needs to be found correctly. Since the GPU on some discrete GFX cards is behind a PCIe switch, checking the _PR3 on downstream port alone is not enough, as the _PR3 can associate to the root port above the PCIe switch. Once the _PR3 is found and BOCO support is correctly marked, use that information to inform the GPU should be reset. This solves an issue that system freeze on a Intel ADL desktop that uses S0ix for sleep and D3cold is supported for the GFX slot. Fixes: 0064b0ce85bb ("drm/amd/pm: enable ASPM by default") Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1885 Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2458 Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c | 3 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 ++++++- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 12 +++++------- 3 files changed, 14 insertions(+), 8 deletions(-)