drm/amdgpu: fix power state when port pm is unavailable
Subsystems (2)
PATCH 1/2 drm/amdgpu: fix power state when port pm is unavailable
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the first year where hybrid graphics platforms using power resources was introduced), but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
---
Hi,
This patch is already three weeks old. One alternative idea was lowering BIOS date in PCI core, but as pcie_port_pm=force did not have the desired effect, I do not think that this would help though.
I have also not contacted linux-pci or Mika about lowering the year due to the lack of a good reason. Might do it later though once ACPICA bug 1333 is sorted out such that lowering the year actually has benefits for a Nvidia laptop (or if some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing 2f5af82eeab2 and b8c9fd5ad4b4 ("track whether if this is a hybrid graphics platform"). There have been some ifdef 0's and reverts in between, so I was not sure if adding the Fixes tag is appropriate.
Kind regards,
Peter
--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +8 -1
@@ -33,6 +33,7 @@ struct amdgpu_atpx { static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;
@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable, +* otherwise the device might fail to power off or power on. +*/ + atpx->functions.power_cntl = !amdgpu_atpx_priv.bridge_pm_usable; atpx->is_hybrid = true; }
@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;
@@ -485,6 +491,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) } amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }
Signature · 6 lines
2.10.2 _______________________________________________ amd-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): 20161123012225.11586-1-peter-VTkQYDcBqhK7DlmcbJSQ7g@public.gmane.org
PATCH 2/2 drm/radeon: fix power state when port pm is unavailable
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the first year where hybrid graphics platforms using power resources was introduced), but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
--- drivers/gpu/drm/radeon/radeon_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/gpu/drm/radeon/radeon_atpx_handler.c +8 -1
@@ -33,6 +33,7 @@ struct radeon_atpx { static struct radeon_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; struct radeon_atpx atpx;
@@ -199,7 +200,11 @@ static int radeon_atpx_validate(struct radeon_atpx *atpx)if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); atpx->functions.power_cntl = false; - atpx->is_hybrid = true; +/* +* Disable legacy PM methods only when pcie port PM is usable, +* otherwise the device might fail to power off or power on. +*/ + atpx->functions.power_cntl = !radeon_atpx_priv.bridge_pm_usable; }return 0 ;
@@ -469,6 +474,7 @@ static int radeon_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool radeon_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;
@@ -482,6 +488,7 @@ static bool radeon_atpx_pci_probe_handle(struct pci_dev *pdev) radeon_atpx_priv.dhandle = dhandle; radeon_atpx_priv.atpx.handle = atpx_handle; + radeon_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }
Signature · 6 lines
2.10.2 _______________________________________________ amd-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Am 23.11.2016 um 02:22 schrieb Peter Wu:
Quoted 5 lines
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Quoted 8 lines
Ideally the BIOS date in the PCI core is lowered to 2013 (the first year where hybrid graphics platforms using power resources was introduced), but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by Nayan Deshmukh <[email protected]>Signed-off-by Peter Wu <[email protected]>
Even when lowering the BIOS date this sounds reasonable to me in the case of pcie_port_pm=off.
But since I'm not very deep into power management the patch is only Acked-by: Christian König <[email protected]>.
Regards,
Christian.
Quoted 57 lines
---
Hi,This patch is already three weeks old. One alternative idea was lowering BIOS date in PCI core, but as pcie_port_pm=force did not have the desired effect, I do not think that this would help though.
I have also not contacted linux-pci or Mika about lowering the year due to the lack of a good reason. Might do it later though once ACPICA bug 1333 is sorted out such that lowering the year actually has benefits for a Nvidia laptop (or if some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andb8c9fd5ad4b4("track whether if this is a hybrid graphics platform"). There have been some ifdef 0's and reverts in between, so I was not sure if adding the Fixes tag is appropriate.Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +7
-1@@ -33,6 +33,7 @@ struct amdgpu_atpx { static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable, +* otherwise the device might fail to power off or power on. +*/ + atpx->functions.power_cntl = !amdgpu_atpx_priv.bridge_pm_usable; atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) · line 497
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0@@ -485,6 +491,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) } amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Alex Deucher
Reviewed-byQuoted 11 lines · Peter Wu
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the first year where hybrid graphics platforms using power resources was introduced), but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Quoted 8 lines
Reported-and-tested-by Nayan Deshmukh <[email protected]>Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was lowering BIOS date in PCI core, but as pcie_port_pm=force did not have the desired effect, I do not think that this would help though.
Thanks for doing this.
Quoted 10 lines
I have also not contacted linux-pci or Mika about lowering the year due to the lack of a good reason. Might do it later though once ACPICA bug 1333 is sorted out such that lowering the year actually has benefits for a Nvidia laptop (or if some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andb8c9fd5ad4b4("track whether if this is a hybrid graphics platform"). There have been some ifdef 0's and reverts in between, so I was not sure if adding the Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't attempt to do d3cold at all. 4.8 and older kernels have:c63695cc5e5f685e924e25a8f9555f6e846f1fc6 (drm/amdgpu: work around lack of upstream ACPI support for D3cold)
which was reverted for 4.9.
Series is:
Quoted 51 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +8
-2@@ -33,6 +33,7 @@ struct amdgpu_atpx { static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable, +* otherwise the device might fail to power off or power on. +*/ + atpx->functions.power_cntl = !amdgpu_atpx_priv.bridge_pm_usable; atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) } amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; } --2.10.2
Quoted 4 lines
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CADnq5_On72XmLrE3wjD0u1=tmwghv6GofzfEGM5mhGoVtJav+Q-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Quoted 34 lines · Alex Deucher
Quoted 11 lines · Peter Wu
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the first year where hybrid graphics platforms using power resources was introduced), but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Quoted 8 lines
Reported-and-tested-by Nayan Deshmukh <[email protected]>Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was lowering BIOS date in PCI core, but as pcie_port_pm=force did not have the desired effect, I do not think that this would help though.
Thanks for doing this.
Quoted 10 lines
I have also not contacted linux-pci or Mika about lowering the year due to the lack of a good reason. Might do it later though once ACPICA bug 1333 is sorted out such that lowering the year actually has benefits for a Nvidia laptop (or if some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andb8c9fd5ad4b4("track whether if this is a hybrid graphics platform"). There have been some ifdef 0's and reverts in between, so I was not sure if adding the Fixes tag is appropriate.
Quoted 5 lines
I don't think we need to cc stable. In kernel 4.8 we don't attempt to do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work around lack of upstream ACPI support for D3cold)
which was reverted for 4.9.
That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grep v4.8 v4.8
Can you double-check? v4.8 was the first kernel with D3cold support in PCI core.
Quoted 2 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>
Thanks!
Kind regards,
Peter
Quoted 49 lines
Quoted 49 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +8
-1@@ -33,6 +33,7 @@ struct amdgpu_atpx { static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable, +* otherwise the device might fail to power off or power on. +*/ + atpx->functions.power_cntl = !amdgpu_atpx_priv.bridge_pm_usable; atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) } amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }
Quoted 2 lines
Quoted 2 lines
--
2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Quoted 58 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm is unavailableQuoted 6 lines · Alex Deucher
Quoted 5 lines · Peter Wu
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Quoted 7 lines
Quoted 7 lines
Ideally the BIOS date in the PCI core is lowered to 2013 (the first year where hybrid graphics platforms using power resources was introduced), but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was lowering
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the desired
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the year due to
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA bug 1333 is
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a Nvidia laptop
(or if
Quoted 4 lines
Quoted 4 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andb8c9fd5ad4b4("track whether if this is a hybrid graphics platform"). Therehave
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure if adding
the
Quoted 5 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't attempt to do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work aroundlack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.
Quoted 7 lines
That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grep v4.8 v4.8Can you double-check? v4.8 was the first kernel with D3cold support in PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes, we should apply this to 4.8.
Alex
Quoted 54 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 9 lines
Review
Quoted 9 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 9 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx)
Quoted 9 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +5
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable, +* otherwise the device might fail to power off or power on. +*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 7 lines
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;
Quoted 12 lines
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): MWHPR12MB16949832E9DB44C53995493AF7B70-Gy0DoCVfaSW4WA4dJ5YXGAdYzm3356FpvxpqHgZTriW3zl9H0oFU5g@public.gmane.org
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
I mistakenly saw the v2 patch and didn't realise it was for radeon not amdgpu - this branch and Linus's tree are currently not booting for me
Quoted 147 lines · Alexander
Quoted 10 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm is unavailableQuoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <[email protected]>
Quoted 6 lines · wrote:
Quoted 6 lines
Quoted 6 lines
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the first
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 11 lines
Quoted 5 lines
Quoted 5 lines
but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was
lowering
Quoted 9 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the desired
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the year
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA bug
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a Nvidia
laptop
Quoted 4 lines
(or if
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure if
adding
Quoted 10 lines
the
Quoted 5 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't attempt to do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work aroundlack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.Quoted 7 lines
That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grep v4.8 v4.8Can you double-check? v4.8 was the first kernel with D3cold support in PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes, we should apply this to 4.8.
Alex
Quoted 24 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 9 lines
Review
Quoted 9 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 14 lines
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable,
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 206
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off or power on.
Quoted 25 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool · line 494
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
parent_pdev->bridge_d3;Quoted 3 lines
Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0return true; }
Quoted 7 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
--
2.10.2_______________________________________________
amd-gfx mailing list [email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CAHbf0-FgRU=6tYe2SYcm8E0ZvJuRFoqPpk77_DZdQKXdxZSsGg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working kernel? I suppose your BIOS date (/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?
Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon not amdgpu - this branch and Linus's tree are currently not booting for me
What do you mean by this?
Kind regards,
Peter
Quoted 146 lines
Quoted 143 lines · Alexander
Quoted 10 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm is unavailableQuoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <[email protected]>
Quoted 6 lines · wrote:
Quoted 6 lines
Quoted 6 lines
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still be marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the first
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 11 lines
Quoted 5 lines
Quoted 5 lines
but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was
lowering
Quoted 9 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the desired
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the year
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA bug
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a Nvidia
laptop
Quoted 4 lines
(or if
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure if
adding
Quoted 10 lines
the
Quoted 5 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't attempt to do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work aroundlack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.Quoted 7 lines
That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grep v4.8 v4.8Can you double-check? v4.8 was the first kernel with D3cold support in PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes, we should apply this to 4.8.
Alex
Quoted 24 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 9 lines
Review
Quoted 9 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 14 lines
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM is usable,
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 206
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off or power on.
Quoted 21 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Hi
The previous working kernel, was the patch before this - reverting fixes it
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't kept and the wifi driver also complains, will see if I can get a slow motion capture of the errors and screen shot them
Regards
Mike
Quoted 149 lines · On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]…
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working kernel? I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon not amdgpu - this branch and Linus's tree are currently not booting for me
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 10 lines
Quoted 8 lines · wrote:
Quoted 8 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm is unavailableQuoted 5 lines
Quoted 5 lines
Quoted 2 lines · Alex Deucher
Quoted 1 line
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <[email protected]>
Quoted 2 lines · wrote:
Quoted 2 lines
Quoted 2 lines
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still
be
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 9 lines
Quoted 9 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 6 lines
Quoted 5 lines
Quoted 5 lines
but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the
year
Quoted 6 lines
Quoted 6 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA bug
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a
Nvidia
Quoted 7 lines
Quoted 7 lines
laptop
Quoted 4 lines
(or if
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsQuoted 8 lines
Quoted 8 lines
platform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure if
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work aroundlack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grepv4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold support
in
Quoted 39 lines
Quoted 39 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes, we should apply this to 4.8.
Alex
Quoted 32 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 9 lines
Review
Quoted 9 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 205
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port
Quoted 33 lines
PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off or
Quoted 28 lines
Quoted 28 lines
power on.
Quoted 27 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool · line 494
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CAHbf0-HfHGt5HvAosxgVf=rgevPK3KqBCQt4cJqjDiEaQWM1pQ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
He's some screen shots of the booting screens, apologies they're not that clear
Quoted 165 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian…
Hi
The previous working kernel, was the patch before this - reverting fixes it
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't kept and the wifi driver also complains, will see if I can get a slow motion capture of the errors and screen shot them
Regards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working kernel? I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon not amdgpu - this branch and Linus's tree are currently not booting for me
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 10 lines
Quoted 8 lines · wrote:
Quoted 8 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm is unavailableQuoted 5 lines
Quoted 5 lines
Quoted 2 lines · Alex Deucher
Quoted 1 line
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <[email protected]>
Quoted 2 lines · wrote:
Quoted 2 lines
Quoted 2 lines
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still
be
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 9 lines
Quoted 9 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 6 lines
Quoted 5 lines
Quoted 5 lines
but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the
year
Quoted 6 lines
Quoted 6 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA bug
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a
Nvidia
Quoted 7 lines
Quoted 7 lines
laptop
Quoted 4 lines
(or if
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsQuoted 8 lines
Quoted 8 lines
platform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure if
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work aroundlack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grepv4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold support
in
Quoted 39 lines
Quoted 39 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes, we should apply this to 4.8.
Alex
Quoted 32 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 9 lines
Review
Quoted 9 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 205
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port
Quoted 34 lines
PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off or
Quoted 28 lines
Quoted 28 lines
power on.
Quoted 27 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool · line 494
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CAHbf0-HHFFthD3Li0wM8x5rciRreeWoE0SkQOvsENT9+ZRFEcg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+ amdgpu_atpx_detect
+ amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+ amdgpu_atpx_init
+ amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usable
I wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1 -0
@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) status =acpi_get_handle (dhandle,"ATPX" , &atpx_handle);if (ACPI_FAILURE (status)) { amdgpu_atpx_priv.other_handle = dhandle; + amdgpu_atpx_priv.bridge_pm_usable = true;return false; } amdgpu_atpx_priv.dhandle = dhandle;
This is not intended for merging, but just testing a hypothesis.
Kind regards,
Peter
Quoted 204 lines · Mike Lothian
He's some screen shots of the booting screens, apologies they're not that clear
Quoted 197 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian <[email protected]> wrote:
Hi
The previous working kernel, was the patch before this - reverting fixes it
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't kept and the wifi driver also complains, will see if I can get a slow motion capture of the errors and screen shot them
Regards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working kernel? I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon not amdgpu - this branch and Linus's tree are currently not booting for me
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 10 lines
Quoted 8 lines · wrote:
Quoted 8 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm is unavailableQuoted 5 lines
Quoted 5 lines
Quoted 2 lines · Alex Deucher
Quoted 1 line
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <[email protected]>
Quoted 2 lines · wrote:
Quoted 2 lines
Quoted 2 lines
When PCIe port PM is not enabled (system BIOS is pre-2015 or the pcie_port_pm=off parameter is set), legacy ATPX PM should still
be
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
marked as supported. Otherwise the GPU can fail to power on after runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 9 lines
Quoted 9 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 6 lines
Quoted 5 lines
Quoted 5 lines
but that seems more risky at this point and would not solve the pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea was
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the
year
Quoted 6 lines
Quoted 6 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA bug
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a
Nvidia
Quoted 7 lines
Quoted 7 lines
laptop
Quoted 4 lines
(or if
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsQuoted 8 lines
Quoted 8 lines
platform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure if
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work aroundlack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f | grepv4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold support
in
Quoted 39 lines
Quoted 39 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes, we should apply this to 4.8.
Alex
Quoted 32 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 9 lines
Review
Quoted 9 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 205
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off or
Quoted 28 lines
Quoted 28 lines
power on.
Quoted 27 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { +struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool · line 494
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Hi
axion fireburn # acpidump > acpidump
Wrong checksum for FADT!
ACPI Dump attached
axion fireburn # lspci -nn
00:00.0 Host bridge [0600]: Intel Corporation Skylake Host Bridge/DRAM Registers [8086:1910] (rev 07)
00:01.0 PCI bridge [0604]: Intel Corporation Skylake PCIe Controller (x16) [8086:1901] (rev 07)
00:02.0 VGA compatible controller [0300]: Intel Corporation HD Graphics 530 [8086:191b] (rev 06)
00:04.0 Signal processing controller [1180]: Intel Corporation Skylake Processor Thermal Subsystem [8086:1903] (rev 07)
00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H USB 3.0 xHCI Controller [8086:a12f] (rev 31)
00:14.2 Signal processing controller [1180]: Intel Corporation Sunrise Point-H Thermal subsystem [8086:a131] (rev 31)
00:16.0 Communication controller [0780]: Intel Corporation Sunrise Point-H CSME HECI #1 [8086:a13a] (rev 31)
00:17.0 SATA controller [0106]: Intel Corporation Sunrise Point-H SATA Controller [AHCI mode] [8086:a103] (rev 31)
00:1c.0 PCI bridge [0604]: Intel Corporation Sunrise Point-H PCI Express Root Port #1 [8086:a110] (rev f1)
00:1c.4 PCI bridge [0604]: Intel Corporation Sunrise Point-H PCI Express Root Port #5 [8086:a114] (rev f1)
00:1c.5 PCI bridge [0604]: Intel Corporation Sunrise Point-H PCI Express Root Port #6 [8086:a115] (rev f1)
00:1c.6 PCI bridge [0604]: Intel Corporation Sunrise Point-H PCI Express Root Port #7 [8086:a116] (rev f1)
00:1d.0 PCI bridge [0604]: Intel Corporation Sunrise Point-H PCI Express Root Port #9 [8086:a118] (rev f1)
00:1f.0 ISA bridge [0601]: Intel Corporation Sunrise Point-H LPC Controller [8086:a14e] (rev 31)
00:1f.2 Memory controller [0580]: Intel Corporation Sunrise Point-H PMC [8086:a121] (rev 31)
00:1f.3 Audio device [0403]: Intel Corporation Sunrise Point-H HD Audio [8086:a170] (rev 31)
00:1f.4 SMBus [0c05]: Intel Corporation Sunrise Point-H SMBus [8086:a123] (rev 31)
01:00.0 Display controller [0380]: Advanced Micro Devices, Inc. [AMD/ATI] Amethyst XT [Radeon R9 M295X] [1002:6921]
3b:00.0 Ethernet controller [0200]: Qualcomm Atheros Killer E2400 Gigabit Ethernet Controller [1969:e0a1] (rev 10)
3c:00.0 Network controller [0280]: Qualcomm Atheros QCA6174 802.11ac Wireless Network Adapter [168c:003e] (rev 32)
3d:00.0 Unassigned class [ff00]: Realtek Semiconductor Co., Ltd. RTS5227 PCI Express Card Reader [10ec:5227] (rev 01)
3e:00.0 Non-Volatile memory controller [0108]: Samsung Electronics Co Ltd NVMe SSD Controller [144d:a802] (rev 01)
I expected an instant lock up when insmodding the modules however that didn't happen
Please also find the dmesg attached
On Sat, 26 Nov 2016 at 11:39 Peter Wu <[email protected]> wrote:
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+ amdgpu_atpx_detect
+ amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+ amdgpu_atpx_init
+ amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usable
I wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
(quoted diff) +0 -0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index 951addf..5c23382 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0 -0
@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev
*pdev)
status = acpi_get_handle(dhandle, "ATPX", &atpx_handle);
if (ACPI_FAILURE(status)) {
amdgpu_atpx_priv.other_handle = dhandle;
+ amdgpu_atpx_priv.bridge_pm_usable = true;
return false;
}
amdgpu_atpx_priv.dhandle = dhandle;This is not intended for merging, but just testing a hypothesis.
Kind regards,
Peter
Quoted 10 lines · Mike Lothian
He's some screen shots of the booting screens, apologies they're not that clear
Quoted 3 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian…
Hi
The previous working kernel, was the patch before this - reverting
fixes it
Quoted 33 lines
Quoted 33 lines
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't kept and the wifi driver also complains, will see if I can get a slow motion capture of the errors and screen shot them
Regards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working kernel? I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon not amdgpu - this branch and Linus's tree are currently not booting for me
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 6 lines · wrote:
Quoted 6 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when port pm
is
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines
unavailable
Quoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <[email protected]>
Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
When PCIe port PM is not enabled (system BIOS is pre-2015 or
the
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
pcie_port_pm=off parameter is set), legacy ATPX PM should
still
Quoted 2 lines
Quoted 2 lines
be
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
marked as supported. Otherwise the GPU can fail to power on
after
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 4 lines
Quoted 4 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 1 line
Quoted 1 line
Quoted 1 line
but that seems more risky at this point and would not solve
the
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 4 lines
Quoted 4 lines
pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea
was
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines
Quoted 3 lines
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
effect, I
Quoted 2 lines
Quoted 1 line
do not think that this would help though.
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering the
year
Quoted 3 lines
Quoted 3 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA
bug
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines
Quoted 3 lines
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a
Nvidia
Quoted 2 lines
Quoted 2 lines
laptop
Quoted 1 line
(or if
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not sure
if
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: work
around
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
lack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f |
grep
Quoted 4 lines
Quoted 4 lines
v4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold
support
Quoted 4 lines
Quoted 4 lines
in
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so yes,
we
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
should apply this to 4.8.
Alex
Quoted 11 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 2 lines
Quoted 2 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9
++++++++-
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 3 lines
Review
Quoted 3 lines
1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 205
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie
port
Quoted 3 lines
Quoted 3 lines
PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power
off
Quoted 4 lines
Quoted 4 lines
or
Quoted 3 lines
Quoted 3 lines
power on.
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0*/ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev
*pdev)
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum · line 479
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0{ + struct pci_dev *parent_pdev =
pci_upstream_bridge(pdev);
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
I can confirm the patch allows my machine to boot again
Quoted 247 lines · On Sat, 26 Nov 2016 at 11:39 Peter Wu <[email protected]…
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+
amdgpu_atpx_detect
+amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+amdgpu_atpx_init
+amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usableI wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) status =acpi_get_handle (dhandle,"ATPX" , &atpx_handle);if (ACPI_FAILURE (status)) { amdgpu_atpx_priv.other_handle = dhandle; + amdgpu_atpx_priv.bridge_pm_usable = true;return false; }amdgpu_atpx_priv.dhandle = dhandle;This is not intended for merging, but just testing a hypothesis.
Kind regards,
PeterQuoted 10 lines · Mike Lothian
He's some screen shots of the booting screens, apologies they're not that clear
Quoted 3 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian…
Hi
The previous working kernel, was the patch before this - reverting
fixes it
Quoted 4 lines
Quoted 4 lines
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't
kept
Quoted 9 lines
Quoted 9 lines
and the wifi driver also complains, will see if I can get a slow motion capture of the errors and screen shot them
Regards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Quoted 7 lines
Quoted 7 lines
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working
kernel?
Quoted 4 lines
Quoted 4 lines
I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 1 line
I mistakenly saw the v2 patch and didn't realise it was for radeon
not
Quoted 1 line
Quoted 1 line
Quoted 1 line
amdgpu - this branch and Linus's tree are currently not booting for
me
Quoted 8 lines
Quoted 8 lines
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 6 lines · wrote:
Quoted 6 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when portpm is
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
unavailable
Quoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
When PCIe port PM is not enabled (system BIOS is pre-2015 or
the
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
pcie_port_pm=off parameter is set), legacy ATPX PM should
still
Quoted 2 lines
Quoted 2 lines
be
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
marked as supported. Otherwise the GPU can fail to power on
after
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 4 lines
Quoted 4 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 1 line
Quoted 1 line
Quoted 1 line
but that seems more risky at this point and would not solve
the
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 4 lines
Quoted 4 lines
pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea
was
Quoted 11 lines
Quoted 11 lines
Quoted 3 lines
Quoted 3 lines
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering
the
Quoted 4 lines
Quoted 4 lines
year
Quoted 3 lines
Quoted 3 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA
bug
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines
Quoted 3 lines
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a
Nvidia
Quoted 2 lines
Quoted 2 lines
laptop
Quoted 1 line
(or if
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not
sure if
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: workaround
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
lack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f |grep
Quoted 4 lines
Quoted 4 lines
v4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold
support
Quoted 4 lines
Quoted 4 lines
in
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so
yes, we
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
should apply this to 4.8.
Alex
Quoted 11 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 2 lines
Quoted 2 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9++++++++-
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 3 lines
Review
Quoted 3 lines
1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 205
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port
Quoted 3 lines
Quoted 3 lines
PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off
Quoted 8 lines
Quoted 8 lines
or
Quoted 7 lines
Quoted 7 lines
power on.
Quoted 6 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 3 lines
Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }
Quoted 24 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0*/ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev*pdev)
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum · line 479
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0{ + struct pci_dev *parent_pdev =
pci_upstream_bridge(pdev);Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CAHbf0-GjBMb2GfFoTADxYjfuLU=+hkJ8AcZCbh70zrpn56cD8A-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Found the problem.
amdgpu_atpx_pci_probe_handle is used to detect the ATPX handle, but this does not necessarily have to be on the dGPU. In fact, on your machine it is located on the Intel iGPU (00:02.0):
[ 134.760509] vga_switcheroo: detected switching method \_SB_.PCI0.GFX0.ATPX handle
The parent device of the iGPU (00:00.0) is the root port which has no PCIe capability, hence bridge_d3 evaluated to false.
Can you try the attached patch?
Kind regards,
Peter
Quoted 268 lines · Mike Lothian
I can confirm the patch allows my machine to boot again
Quoted 264 lines · On Sat, 26 Nov 2016 at 11:39 Peter Wu <[email protected]…
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+
amdgpu_atpx_detect
+amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+amdgpu_atpx_init
+amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usableI wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) status =acpi_get_handle (dhandle,"ATPX" , &atpx_handle);if (ACPI_FAILURE (status)) { amdgpu_atpx_priv.other_handle = dhandle; + amdgpu_atpx_priv.bridge_pm_usable = true;return false; }amdgpu_atpx_priv.dhandle = dhandle;This is not intended for merging, but just testing a hypothesis.
Kind regards,
PeterQuoted 10 lines · Mike Lothian
He's some screen shots of the booting screens, apologies they're not that clear
Quoted 3 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian…
Hi
The previous working kernel, was the patch before this - reverting
fixes it
Quoted 4 lines
Quoted 4 lines
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't
kept
Quoted 9 lines
Quoted 9 lines
and the wifi driver also complains, will see if I can get a slow motion capture of the errors and screen shot them
Regards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Quoted 7 lines
Quoted 7 lines
Hi Mike,
Quoted 2 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working
kernel?
Quoted 4 lines
Quoted 4 lines
I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 1 line
I mistakenly saw the v2 patch and didn't realise it was for radeon
not
Quoted 1 line
Quoted 1 line
Quoted 1 line
amdgpu - this branch and Linus's tree are currently not booting for
me
Quoted 8 lines
Quoted 8 lines
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 6 lines · wrote:
Quoted 6 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when portpm is
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
unavailable
Quoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
When PCIe port PM is not enabled (system BIOS is pre-2015 or
the
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
pcie_port_pm=off parameter is set), legacy ATPX PM should
still
Quoted 2 lines
Quoted 2 lines
be
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
marked as supported. Otherwise the GPU can fail to power on
after
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 4 lines
Quoted 4 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 1 line
Quoted 1 line
Quoted 1 line
but that seems more risky at this point and would not solve
the
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 4 lines
Quoted 4 lines
pcie_port_pm=off issue.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea
was
Quoted 11 lines
Quoted 11 lines
Quoted 3 lines
Quoted 3 lines
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about lowering
the
Quoted 4 lines
Quoted 4 lines
year
Quoted 3 lines
Quoted 3 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once ACPICA
bug
Quoted 6 lines
Quoted 6 lines
Quoted 3 lines
Quoted 3 lines
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits for a
Nvidia
Quoted 2 lines
Quoted 2 lines
laptop
Quoted 1 line
(or if
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not
sure if
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: workaround
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
lack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f |grep
Quoted 4 lines
Quoted 4 lines
v4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold
support
Quoted 4 lines
Quoted 4 lines
in
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so
yes, we
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
should apply this to 4.8.
Alex
Quoted 11 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 2 lines
Quoted 2 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9++++++++-
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 17 lines
Quoted 3 lines
Review
Quoted 3 lines
1 file changed, 8 insertions(+), 1 deletion(-)(quoted diff) +0
-0
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static int amdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_validate(struct · line 205
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie port
Quoted 3 lines
Quoted 3 lines
PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power off
Quoted 4 lines
Quoted 4 lines
or
Quoted 3 lines
Quoted 3 lines
power on.
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static int amdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0*/ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev*pdev)
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int amdgpu_atpx_power_state(enum · line 479
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0{ + struct pci_dev *parent_pdev =
pci_upstream_bridge(pdev);Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0acpi_handle dhandle, atpx_handle;acpi_status status;@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev &&
Quoted 7 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
From
483607ff5c5dc6932e2f854d205cf41f9fbf26c8Mon Sep 17 00:00:00 2001
From: Peter Wu <[email protected]>
Date: Sat, 26 Nov 2016 15:05:01 +0100
Subject: [PATCH] drm/amdgpu: fix check for port PM availability
The ATPX method does not always exist on the dGPU, it may be located at the iGPU. The parent device of the iGPU is the root port for which bridge_d3 is false. This accidentally enables the legacy PM method which conflicts with port PM and prevented the dGPU from powering on.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +9 -2
@@ -476,7 +476,6 @@ static int amdgpu_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool amdgpu_atpx_pci_probe_handle (struct pci_dev *pdev) { -struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;
@@ -491,7 +490,6 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) } amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; - amdgpu_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }
@@ -553,17 +551,25 @@ static bool amdgpu_atpx_detect(void)struct pci_dev *pdev =NULL ;bool has_atpx = false;int vga_count =0 ; +bool d3_supported = false; +struct pci_dev *parent_pdev;while ((pdev =pci_get_class (PCI_CLASS_DISPLAY_VGA <<8 , pdev)) !=NULL ) { vga_count++; has_atpx |= (amdgpu_atpx_pci_probe_handle (pdev) == true); + + parent_pdev =pci_upstream_bridge (pdev); + d3_supported |= parent_pdev && parent_pdev->bridge_d3; }while ((pdev =pci_get_class (PCI_CLASS_DISPLAY_OTHER <<8 , pdev)) !=NULL ) { vga_count++; has_atpx |= (amdgpu_atpx_pci_probe_handle (pdev) == true); + + parent_pdev =pci_upstream_bridge (pdev); + d3_supported |= parent_pdev && parent_pdev->bridge_d3; }if (has_atpx && vga_count ==2 ) {
@@ -571,6 +577,7 @@ static bool amdgpu_atpx_detect(void)printk (KERN_INFO "vga_switcheroo: detected switching method %s handle\n" , acpi_method_name); amdgpu_atpx_priv.atpx_detected = true; + amdgpu_atpx_priv.bridge_pm_usable = d3_supported;amdgpu_atpx_init ();return true; }
Signature · 1 line
2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
The attached patch allows my machine to boot also
Thanks
Mike
Quoted 303 lines · On Sat, 26 Nov 2016 at 14:14 Peter Wu <[email protected]…
Found the problem.
amdgpu_atpx_pci_probe_handleis used to detect the ATPX handle, but this does not necessarily have to be on the dGPU. In fact, on your machine it is located on the Intel iGPU (00:02.0):[ 134.760509] vga_switcheroo: detected switching method\_SB_.PCI0.GFX0.ATPX handle
The parent device of the iGPU (00:00.0) is the root port which has no PCIe capability, hence
bridge_d3evaluated to false.Can you try the attached patch?
Kind regards,
PeterQuoted 47 lines · Mike Lothian
I can confirm the patch allows my machine to boot again
Quoted 43 lines · On Sat, 26 Nov 2016 at 11:39 Peter Wu <[email protected]…
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+
amdgpu_atpx_detect
+amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+amdgpu_atpx_init
+amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usableI wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) status =acpi_get_handle (dhandle,"ATPX" , &atpx_handle);if (ACPI_FAILURE (status)) { amdgpu_atpx_priv.other_handle = dhandle; + amdgpu_atpx_priv.bridge_pm_usable = true;return false; }amdgpu_atpx_priv.dhandle = dhandle;This is not intended for merging, but just testing a hypothesis.
Kind regards,
PeterQuoted 1 line · Mike Lothian
He's some screen shots of the booting screens, apologies they're not
that
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
clear
On Sat, 26 Nov 2016 at 11:00 Mike Lothian <[email protected]>
Quoted 11 lines · wrote:
Quoted 11 lines
Quoted 4 lines
Quoted 4 lines
Hi
The previous working kernel, was the patch before this - reverting
fixes it
Quoted 4 lines
Quoted 4 lines
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't
kept
Quoted 1 line
Quoted 1 line
and the wifi driver also complains, will see if I can get a slow
motion
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
capture of the errors and screen shot them
Regards
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Hi Mike,
Quoted 1 line · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3
error
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
messages and atom bios stuck messages
Can you post a dmesg and acpidump? What was the previous working
kernel?
Quoted 4 lines
Quoted 4 lines
I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015? Does booting with runpm=0 work?Quoted 1 line
I mistakenly saw the v2 patch and didn't realise it was for
radeon
Quoted 2 lines
Quoted 2 lines
not
Quoted 1 line
Quoted 1 line
Quoted 1 line
amdgpu - this branch and Linus's tree are currently not booting
for
Quoted 9 lines
Quoted 9 lines
me
Quoted 8 lines
Quoted 8 lines
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 6 lines · wrote:
Quoted 6 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state whenport
Quoted 8 lines
Quoted 8 lines
pm is
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
unavailable
Quoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
When PCIe port PM is not enabled (system BIOS is
pre-2015 or
Quoted 5 lines
Quoted 5 lines
the
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
pcie_port_pm=off parameter is set), legacy ATPX PM should
still
Quoted 2 lines
Quoted 2 lines
be
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
marked as supported. Otherwise the GPU can fail to power
on
Quoted 4 lines
Quoted 4 lines
after
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013
(the
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
first
Quoted 4 lines
Quoted 4 lines
year
Quoted 1 line
Quoted 1 line
Quoted 1 line
where hybrid graphics platforms using power resources was
introduced),
Quoted 1 line
Quoted 1 line
Quoted 1 line
but that seems more risky at this point and would not
solve
Quoted 4 lines
Quoted 4 lines
the
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
pcie_port_pm=off issue.
Bugzilla:
https://bugs.freedesktop.org/show_bug.cgi?id=98505
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 1 line
Quoted 1 line
Reported-and-tested-by: Nayan Deshmukh
Quoted 5 lines
Quoted 5 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative
idea
Quoted 4 lines
Quoted 4 lines
was
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
lowering
Quoted 2 lines
BIOS
Quoted 1 line
Quoted 1 line
date in PCI core, but as pcie_port_pm=force did not have
the
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
desired
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
effect, I
Quoted 6 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 2 lines
I have also not contacted linux-pci or Mika about
lowering
Quoted 5 lines
Quoted 5 lines
the
Quoted 4 lines
Quoted 4 lines
year
Quoted 3 lines
Quoted 3 lines
due to
Quoted 2 lines
the
Quoted 1 line
Quoted 1 line
lack of a good reason. Might do it later though once
ACPICA
Quoted 4 lines
Quoted 4 lines
bug
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
1333 is
Quoted 2 lines
sorted
Quoted 1 line
Quoted 1 line
out such that lowering the year actually has benefits
for a
Quoted 20 lines
Quoted 20 lines
Quoted 11 lines
Quoted 11 lines
Nvidia
Quoted 10 lines
Quoted 10 lines
laptop
Quoted 4 lines
(or if
Quoted 3 lines
Quoted 3 lines
some amdgpu problem can be solved by this).
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 1 line
Quoted 1 line
Quoted 1 line
b8c9fd5ad4b4("track whether if this is a hybrid graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not
sure if
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
adding
Quoted 4 lines
the
Quoted 3 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: workQuoted 8 lines
Quoted 8 lines
around
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
lack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains
c39b487f195b93235ee76384427467786f7bf29f|Quoted 26 lines
Quoted 26 lines
grep
Quoted 4 lines
Quoted 4 lines
v4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold
support
Quoted 4 lines
Quoted 4 lines
in
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so
yes, we
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
should apply this to 4.8.
Alex
Quoted 11 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 2 lines
Quoted 2 lines
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9++++++++-
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
1 file changed, 8 insertions(+), 1 deletion(-)diff --git
a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int
Quoted 7 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static intamdgpu_atpx_validate(struct
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
amdgpu_atpx*atpx)Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1atpx->is_hybrid = false;if (valid_bits &ATPX_MS_HYBRID_GFX_SUPPORTED ) {printk ("ATPX Hybrid Graphics\n" ); - atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie
Quoted 4 lines
Quoted 4 lines
port
Quoted 3 lines
Quoted 3 lines
PM
Quoted 2 lines
Quoted 2 lines
is usable,
Quoted 1 line
Quoted 1 line
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0+* otherwise the device might fail to power
Quoted 10 lines
Quoted 10 lines
off
Quoted 9 lines
Quoted 9 lines
or
Quoted 8 lines
Quoted 8 lines
power on.
Quoted 7 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int · line 208
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0atpx->is_hybrid = true; }@@ -471,6 +476,7 @@ static intamdgpu_atpx_power_state(enum
Quoted 10 lines
Quoted 10 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
vga_switcheroo_client_idid,Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0*/ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev*pdev)
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static int · line 479
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0{ + struct pci_dev *parent_pdev =
pci_upstream_bridge(pdev);Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0acpi_handle dhandle, atpx_handle;acpi_status status;
Quoted 14 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 1 line
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev&&
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-1return true; } --2.10.2
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CAHbf0-HzBPXOqS_XLctTu-eTuxUUwP62_TcODVPwKb8T2PFrFg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Quoted 21 lines · Mike Lothian
The attached patch allows my machine to boot also
Thanks
Mike
Quoted 14 lines · On Sat, 26 Nov 2016 at 14:14 Peter Wu <[email protected]…
Found the problem.
amdgpu_atpx_pci_probe_handleis used to detect the ATPX handle, but this does not necessarily have to be on the dGPU. In fact, on your machine it is located on the Intel iGPU (00:02.0):[ 134.760509] vga_switcheroo: detected switching method\_SB_.PCI0.GFX0.ATPX handle
The parent device of the iGPU (00:00.0) is the root port which has no PCIe capability, hence
bridge_d3evaluated to false.Can you try the attached patch?
Looks good to me. Patch for radeon included as well.
Alex
Quoted 304 lines
Quoted 304 lines
Kind regards,
PeterQuoted 299 lines · Mike Lothian
I can confirm the patch allows my machine to boot again
Quoted 295 lines · On Sat, 26 Nov 2016 at 11:39 Peter Wu <[email protected]…
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+
amdgpu_atpx_detect
+amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+amdgpu_atpx_init
+amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usableI wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) status =acpi_get_handle (dhandle,"ATPX" , &atpx_handle);if (ACPI_FAILURE (status)) { amdgpu_atpx_priv.other_handle = dhandle; + amdgpu_atpx_priv.bridge_pm_usable = true;return false; }amdgpu_atpx_priv.dhandle = dhandle;This is not intended for merging, but just testing a hypothesis.
Kind regards,
PeterQuoted 4 lines · Mike Lothian
He's some screen shots of the booting screens, apologies they're not that
clearQuoted 8 lines
Quoted 3 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian…
Hi
The previous working kernel, was the patch before this - reverting
fixes it
Quoted 5 lines
Quoted 5 lines
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't
kept
Quoted 10 lines
Quoted 10 lines
and the wifi driver also complains, will see if I can get a slow motion
capture of the errors and screen shot themRegards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Quoted 8 lines
Quoted 8 lines
Hi Mike,
Quoted 3 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error
messages and atom bios stuck messagesCan you post a dmesg and acpidump? What was the previous working
kernel?
Quoted 6 lines
Quoted 6 lines
I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015?
Does booting with runpm=0 work?Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon
not
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
amdgpu - this branch and Linus's tree are currently not booting for
me
Quoted 10 lines
Quoted 10 lines
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 2 lines
wrote:
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when portpm is
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
unavailable
Quoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 2 lines · wrote:
Quoted 2 lines
Quoted 2 lines
When PCIe port PM is not enabled (system BIOS is
pre-2015 orthe
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
pcie_port_pm=off parameter is set), legacy ATPX PM should
still
Quoted 3 lines
Quoted 3 lines
be
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
marked as supported. Otherwise the GPU can fail to power on
after
Quoted 11 lines
Quoted 11 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 6 lines
Quoted 6 lines
year
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
where hybrid graphics platforms using power resources was
introduced),
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
but that seems more risky at this point and would not solve
the
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
pcie_port_pm=off issue.
Bugzilla:
https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 6 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea
was
Quoted 5 lines
Quoted 5 lines
Quoted 4 lines
Quoted 4 lines
lowering
Quoted 3 lines
BIOS
Quoted 2 lines
Quoted 2 lines
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
effect, I
Quoted 7 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 3 lines
I have also not contacted linux-pci or Mika about
loweringthe
Quoted 5 lines
Quoted 5 lines
year
Quoted 4 lines
Quoted 4 lines
due to
Quoted 3 lines
the
Quoted 2 lines
Quoted 2 lines
lack of a good reason. Might do it later though once ACPICA
bug
Quoted 9 lines
Quoted 9 lines
Quoted 4 lines
Quoted 4 lines
1333 is
Quoted 3 lines
sorted
Quoted 2 lines
Quoted 2 lines
out such that lowering the year actually has benefits for a
Nvidia
Quoted 4 lines
Quoted 4 lines
laptop
Quoted 3 lines
(or if
Quoted 2 lines
Quoted 2 lines
some amdgpu problem can be solved by this).
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 2 lines
Quoted 2 lines
Quoted 2 lines
b8c9fd5ad4b4("track whether if this is a hybrid
graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not
sure if
Quoted 9 lines
Quoted 9 lines
Quoted 6 lines
Quoted 6 lines
adding
Quoted 5 lines
the
Quoted 4 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: workaround
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
lack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f |grep
Quoted 4 lines
Quoted 4 lines
v4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold
support
Quoted 4 lines
Quoted 4 lines
in
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so
yes, we
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
should apply this to 4.8.
Alex
Quoted 11 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 2 lines
Quoted 2 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9++++++++-
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 4 lines
Quoted 4 lines
1 file changed, 8 insertions(+), 1 deletion(-)diff --git
a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 8 lines
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static intamdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
atpx->is_hybrid = false; if (valid_bits & ATPX_MS_HYBRID_GFX_SUPPORTED) { printk("ATPX Hybrid Graphics\n");drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1- atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie
port
Quoted 4 lines
Quoted 4 lines
PM
Quoted 3 lines
Quoted 3 lines
is usable,
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
+ * otherwise the device might fail to
poweroff
Quoted 5 lines
Quoted 5 lines
or
Quoted 4 lines
Quoted 4 lines
power on.
Quoted 3 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
atpx->is_hybrid = true; }drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -471,6 +476,7 @@ static intamdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 2 lines
Quoted 2 lines
*/
static bool amdgpu_atpx_pci_probe_handle(structpci_dev*pdev)
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
{
+ structpci_dev*parent_pdev=
pci_upstream_bridge(pdev);Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 4 lines
acpi_handle dhandle, atpx_handle; acpi_status status;drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev&&
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
return true;}
--
2.10.2
From cd21b5055cca49b30b0caaf1107a9aaeb60a447f Mon Sep 17 00:00:00 2001
From: Alex Deucher <[email protected]>
Date: Mon, 28 Nov 2016 17:23:40 -0500
Subject: [PATCH] drm/radeon: fix check for port PM availability
The ATPX method does not always exist on the dGPU, it may be located at the iGPU. The parent device of the iGPU is the root port for which bridge_d3 is false. This accidentally enables the legacy PM method which conflicts with port PM and prevented the dGPU from powering on.
Ported from amdgpu commit:
drm/amdgpu: fix check for port PM availability
from Peter Wu.
--- drivers/gpu/drm/radeon/radeon_atpx_handler.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)
drivers/gpu/drm/radeon/radeon_atpx_handler.c +9 -2
@@ -479,7 +479,6 @@ static int radeon_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool radeon_atpx_pci_probe_handle (struct pci_dev *pdev) { -struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;
@@ -493,7 +492,6 @@ static bool radeon_atpx_pci_probe_handle(struct pci_dev *pdev) radeon_atpx_priv.dhandle = dhandle; radeon_atpx_priv.atpx.handle = atpx_handle; - radeon_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }
@@ -555,11 +553,16 @@ static bool radeon_atpx_detect(void)struct pci_dev *pdev =NULL ;bool has_atpx = false;int vga_count =0 ; +bool d3_supported = false; +struct pci_dev *parent_pdev;while ((pdev =pci_get_class (PCI_CLASS_DISPLAY_VGA <<8 , pdev)) !=NULL ) { vga_count++; has_atpx |= (radeon_atpx_pci_probe_handle (pdev) == true); + + parent_pdev =pci_upstream_bridge (pdev); + d3_supported |= parent_pdev && parent_pdev->bridge_d3; }/* some newer PX laptops mark the dGPU as a non-VGA display device */
@@ -567,6 +570,9 @@ static bool radeon_atpx_detect(void) vga_count++; has_atpx |= (radeon_atpx_pci_probe_handle (pdev) == true); + + parent_pdev =pci_upstream_bridge (pdev); + d3_supported |= parent_pdev && parent_pdev->bridge_d3; }if (has_atpx && vga_count ==2 ) {
@@ -574,6 +580,7 @@ static bool radeon_atpx_detect(void)printk (KERN_INFO "vga_switcheroo: detected switching method %s handle\n" , acpi_method_name); radeon_atpx_priv.atpx_detected = true; + radeon_atpx_priv.bridge_pm_usable = d3_supported;radeon_atpx_init ();return true; }
Signature · 1 line
2.5.5
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
A message that is not in the archive (sent off-list or not yet received): CADnq5_MfLP+Ym+beD+Oak+CTxvyMVYfb9+DQDZuR9pbBWqwQtA-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Quoted 18 lines · Alex Deucher
Quoted 17 lines · Mike Lothian
The attached patch allows my machine to boot also
Thanks
Mike
Quoted 10 lines · On Sat, 26 Nov 2016 at 14:14 Peter Wu <[email protected]> wrote:
Found the problem.
amdgpu_atpx_pci_probe_handleis used to detect the ATPX handle, but this does not necessarily have to be on the dGPU. In fact, on your machine it is located on the Intel iGPU (00:02.0):[ 134.760509] vga_switcheroo: detected switching method\_SB_.PCI0.GFX0.ATPX handle
Quoted 8 lines
Quoted 4 lines
Quoted 4 lines
The parent device of the iGPU (00:00.0) is the root port which has no PCIe capability, hence
bridge_d3evaluated to false.Can you try the attached patch?
Looks good to me. Patch for radeon included as well.
Alex
Thanks, the port looks good to me. (If still possible, I think you can change Reported-by also to Reported-and-tested-by for the amdgpu patch.)
Kind regards,
Peter
Quoted 391 lines
Quoted 304 lines
Quoted 304 lines
Kind regards,
PeterQuoted 299 lines · Mike Lothian
I can confirm the patch allows my machine to boot again
Quoted 295 lines · On Sat, 26 Nov 2016 at 11:39 Peter Wu <[email protected]> wrote:
Hi Mike,
To grab console output, you could also try to blacklist amdgpu at startup and load it later.
Can you provide your acpidump (and lspci -nn, especially if you do not have a dmesg).
This is the call chain:
+
amdgpu_atpx_detect
+amdgpu_atpx_pci_probe_handle
+ return if ATPX ACPI handle cannot be found
+ set amdgpu_atpx_priv.bridge_pm_usable
+amdgpu_atpx_init
+amdgpu_atpx_validate
+ use amdgpu_atpx_priv.bridge_pm_usableI wonder whether newer devices remove ATPX completely. If that is the case (please provide acpidump!), can you try this hack:
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0@@ -487,6 +487,7 @@ static bool amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev) status =acpi_get_handle (dhandle,"ATPX" , &atpx_handle);if (ACPI_FAILURE (status)) { amdgpu_atpx_priv.other_handle = dhandle; + amdgpu_atpx_priv.bridge_pm_usable = true;return false; }amdgpu_atpx_priv.dhandle = dhandle;This is not intended for merging, but just testing a hypothesis.
Kind regards,
PeterQuoted 4 lines · Mike Lothian
He's some screen shots of the booting screens, apologies they're not that
clearQuoted 8 lines
Quoted 3 lines · On Sat, 26 Nov 2016 at 11:00 Mike Lothian <[email protected]> wrote:
Hi
The previous working kernel, was the patch before this - reverting
fixes it
Quoted 5 lines
Quoted 5 lines
My bios date is quite new 08/05/2016
Most of the messages fly straight off the screen, the journal isn't
kept
Quoted 10 lines
Quoted 10 lines
and the wifi driver also complains, will see if I can get a slow motion
capture of the errors and screen shot themRegards
Mike
On Sat, 26 Nov 2016 at 10:20 Peter Wu <[email protected]> wrote:
Quoted 8 lines
Quoted 8 lines
Hi Mike,
Quoted 3 lines · Mike Lothian
This patch is preventing my laptop from booting, I'm getting D3 error
messages and atom bios stuck messagesCan you post a dmesg and acpidump? What was the previous working
kernel?
Quoted 6 lines
Quoted 6 lines
I suppose your BIOS date (
/sys/class/dmi/id/bios_date) is pre-2015?
Does booting with runpm=0 work?Quoted 2 lines
I mistakenly saw the v2 patch and didn't realise it was for radeon
not
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
amdgpu - this branch and Linus's tree are currently not booting for
me
Quoted 10 lines
Quoted 10 lines
What do you mean by this?
Kind regards,
PeterQuoted 1 line
On Wed, 23 Nov 2016 at 17:16 Deucher, Alexander <
Quoted 2 lines
wrote:
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
-----Original Message-----
From: Peter Wu [mailto:[email protected]]
Sent: Wednesday, November 23, 2016 12:13 PM
To: Alex Deucher
Cc: amd-gfx list; Deucher, Alexander; Nayan Deshmukh
Subject: Re: [PATCH 1/2] drm/amdgpu: fix power state when portpm is
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
unavailable
Quoted 1 line · Alex Deucher
On Tue, Nov 22, 2016 at 8:22 PM, Peter Wu <
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 2 lines · wrote:
Quoted 2 lines
Quoted 2 lines
When PCIe port PM is not enabled (system BIOS is
pre-2015 orthe
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
pcie_port_pm=off parameter is set), legacy ATPX PM should
still
Quoted 3 lines
Quoted 3 lines
be
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
marked as supported. Otherwise the GPU can fail to power on
after
Quoted 11 lines
Quoted 11 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
runtime suspend. This affected a Dell Inspiron 5548.
Ideally the BIOS date in the PCI core is lowered to 2013 (the
first
Quoted 6 lines
Quoted 6 lines
year
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
where hybrid graphics platforms using power resources was
introduced),
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
but that seems more risky at this point and would not solve
the
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
pcie_port_pm=off issue.
Bugzilla:
https://bugs.freedesktop.org/show_bug.cgi?id=98505
Reported-and-tested-by: Nayan DeshmukhQuoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 6 lines
Signed-off-by Peter Wu <[email protected]>---
Hi,This patch is already three weeks old. One alternative idea
was
Quoted 5 lines
Quoted 5 lines
Quoted 4 lines
Quoted 4 lines
lowering
Quoted 3 lines
BIOS
Quoted 2 lines
Quoted 2 lines
date in PCI core, but as pcie_port_pm=force did not have the
desired
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
effect, I
Quoted 7 lines
Quoted 1 line
do not think that this would help though.
Thanks for doing this.
Quoted 3 lines
I have also not contacted linux-pci or Mika about
loweringthe
Quoted 5 lines
Quoted 5 lines
year
Quoted 4 lines
Quoted 4 lines
due to
Quoted 3 lines
the
Quoted 2 lines
Quoted 2 lines
lack of a good reason. Might do it later though once ACPICA
bug
Quoted 9 lines
Quoted 9 lines
Quoted 4 lines
Quoted 4 lines
1333 is
Quoted 3 lines
sorted
Quoted 2 lines
Quoted 2 lines
out such that lowering the year actually has benefits for a
Nvidia
Quoted 4 lines
Quoted 4 lines
laptop
Quoted 3 lines
(or if
Quoted 2 lines
Quoted 2 lines
some amdgpu problem can be solved by this).
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Both patches should probably be Cc stable (4.8+), fixing
2f5af82eeab2andQuoted 2 lines
Quoted 2 lines
Quoted 2 lines
b8c9fd5ad4b4("track whether if this is a hybrid
graphicsplatform"). There
Quoted 2 lines
have
Quoted 1 line
Quoted 1 line
been some ifdef 0's and reverts in between, so I was not
sure if
Quoted 9 lines
Quoted 9 lines
Quoted 6 lines
Quoted 6 lines
adding
Quoted 5 lines
the
Quoted 4 lines
Quoted 1 line
Fixes tag is appropriate.
I don't think we need to cc stable. In kernel 4.8 we don't
attempt to
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
do d3cold at all. 4.8 and older kernels have:
c63695cc5e5f685e924e25a8f9555f6e846f1fc6(drm/amdgpu: workaround
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
lack
Quoted 2 lines
of upstream ACPI support for D3cold)
which was reverted for 4.9.That patch was already reverted in 4.8 as far as I can see:
$ git tag --contains c39b487f195b93235ee76384427467786f7bf29f |grep
Quoted 4 lines
Quoted 4 lines
v4.8
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
v4.8
Can you double-check? v4.8 was the first kernel with D3cold
support
Quoted 4 lines
Quoted 4 lines
in
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
PCI core.
You are right. I was thinking d3cold went upstream in 4.9, so
yes, we
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
should apply this to 4.8.
Alex
Quoted 11 lines
Quoted 3 lines
Series is:
Reviewed-by Alex Deucher <[email protected]>Thanks!
Kind regards,
PeterQuoted 2 lines
Quoted 2 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Kind regards,
Peter--- drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c | 9++++++++-
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines
Quoted 4 lines
Quoted 4 lines
1 file changed, 8 insertions(+), 1 deletion(-)diff --git
a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.cQuoted 5 lines
Quoted 5 lines
index 10b5ddf..951addf 100644
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -33,6 +33,7 @@ struct amdgpu_atpx {Quoted 8 lines
Quoted 8 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0static struct amdgpu_atpx_priv { bool atpx_detected; + bool bridge_pm_usable; /* handle for device - and atpx */ acpi_handle dhandle; acpi_handle other_handle;@@ -200,7 +201,11 @@ static intamdgpu_atpx_validate(struct
amdgpu_atpx*atpx)Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
atpx->is_hybrid = false; if (valid_bits & ATPX_MS_HYBRID_GFX_SUPPORTED) { printk("ATPX Hybrid Graphics\n");drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-1- atpx->functions.power_cntl = false; +/* +* Disable legacy PM methods only when pcie
port
Quoted 4 lines
Quoted 4 lines
PM
Quoted 3 lines
Quoted 3 lines
is usable,
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
+ * otherwise the device might fail to
poweroff
Quoted 5 lines
Quoted 5 lines
or
Quoted 4 lines
Quoted 4 lines
power on.
Quoted 3 lines
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
Quoted 2 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +2
-0+*/ + atpx->functions.power_cntl =!amdgpu_atpx_priv.bridge_pm_usable;
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
atpx->is_hybrid = true; }drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -471,6 +476,7 @@ static intamdgpu_atpx_power_state(enum
vga_switcheroo_client_idid,Quoted 2 lines
Quoted 2 lines
*/
static bool amdgpu_atpx_pci_probe_handle(structpci_dev*pdev)
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
{
+ structpci_dev*parent_pdev=
pci_upstream_bridge(pdev);Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 4 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c static bool
Quoted 4 lines
acpi_handle dhandle, atpx_handle; acpi_status status;drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +0
-0@@ -485,6 +491,7 @@ static bool
amdgpu_atpx_pci_probe_handle(struct pci_dev *pdev)Quoted 5 lines
Quoted 5 lines
drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +1
-0} amdgpu_atpx_priv.dhandle = dhandle; amdgpu_atpx_priv.atpx.handle = atpx_handle; + amdgpu_atpx_priv.bridge_pm_usable = parent_pdev&&
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
parent_pdev->bridge_d3;Quoted 5 lines
Quoted 5 lines
return true;}
--
2.10.2From
cd21b5055cca49b30b0caaf1107a9aaeb60a447fMon Sep 17 00:00:00 2001
From: Alex Deucher <[email protected]>
Date: Mon, 28 Nov 2016 17:23:40 -0500
Subject: [PATCH] drm/radeon: fix check for port PM availabilityThe ATPX method does not always exist on the dGPU, it may be located at the iGPU. The parent device of the iGPU is the root port for which
bridge_d3is false. This accidentally enables the legacy PM method which conflicts with port PM and prevented the dGPU from powering on.Ported from amdgpu commit:
drm/amdgpu: fix check for port PM availability
from Peter Wu.Fixes d3ac31f3b4bf9fad (drm/radeon: fix power state when port pm is unavailable (v2))Signed-off-by Alex Deucher <[email protected]>Cc Peter Wu <[email protected]>Cc <[email protected]> # 4.8+--- drivers/gpu/drm/radeon/radeon_atpx_handler.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)drivers/gpu/drm/radeon/radeon_atpx_handler.c +9
-2@@ -479,7 +479,6 @@ static int radeon_atpx_power_state(enum vga_switcheroo_client_id id,*/ static bool radeon_atpx_pci_probe_handle (struct pci_dev *pdev) { -struct pci_dev *parent_pdev =pci_upstream_bridge (pdev);acpi_handle dhandle, atpx_handle;acpi_status status;@@ -493,7 +492,6 @@ static bool radeon_atpx_pci_probe_handle(struct pci_dev *pdev) radeon_atpx_priv.dhandle = dhandle; radeon_atpx_priv.atpx.handle = atpx_handle; - radeon_atpx_priv.bridge_pm_usable = parent_pdev && parent_pdev->bridge_d3;return true; }@@ -555,11 +553,16 @@ static bool radeon_atpx_detect(void)struct pci_dev *pdev =NULL ;bool has_atpx = false;int vga_count =0 ; +bool d3_supported = false; +struct pci_dev *parent_pdev;while ((pdev =pci_get_class (PCI_CLASS_DISPLAY_VGA <<8 , pdev)) !=NULL ) { vga_count++; has_atpx |= (radeon_atpx_pci_probe_handle (pdev) == true); + + parent_pdev =pci_upstream_bridge (pdev); + d3_supported |= parent_pdev && parent_pdev->bridge_d3; }/* some newer PX laptops mark the dGPU as a non-VGA display device */ @@ -567,6 +570,9 @@ static bool radeon_atpx_detect(void) vga_count++; has_atpx |= (radeon_atpx_pci_probe_handle (pdev) == true); + + parent_pdev =pci_upstream_bridge (pdev); + d3_supported |= parent_pdev && parent_pdev->bridge_d3; }if (has_atpx && vga_count ==2 ) {@@ -574,6 +580,7 @@ static bool radeon_atpx_detect(void)printk (KERN_INFO "vga_switcheroo: detected switching method %s handle\n" , acpi_method_name); radeon_atpx_priv.atpx_detected = true; + radeon_atpx_priv.bridge_pm_usable = d3_supported;radeon_atpx_init ();return true; }Signature · 1 line
2.5.5
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx