All of lore.kernel.org
 help / color / mirror / Atom feed
* [PULL] amdkfd-fixes
@ 2014-12-30 19:35 Oded Gabbay
  0 siblings, 0 replies; 5+ messages in thread
From: Oded Gabbay @ 2014-12-30 19:35 UTC (permalink / raw)
  To: airlied, dri-devel; +Cc: alexander.deucher

Hi Dave,

A couple of bug fixes for 3.19-rc3, or for -rc4 if you decide to skip -rc3.

Highlights:

- Link order changes in drm/Makefile and drivers/Makefile to fix issue
  when amdkfd, radeon and amd_iommu_v2 are compiled inside the kernel 
  image.

- Consider kernel configuration (using #IFDEFs) when radeon initializes
  amdkfd, due to a specific configuration that makes symbol_request() 
  return a non-NULL value when a symbol doesn't exists. Rusty Russel
  is helping me to find the root cause, but it may take a while because
  of year-end so I'm sending this as a band-aid solution.
  
Other fixes include topology, pasid allocations and queues accounting.

Thanks and a happy new year

	Oded

The following changes since commit b7392d2247cfe6771f95d256374f1a8e6a6f48d6:

  Linux 3.19-rc2 (2014-12-28 16:49:37 -0800)

are available in the git repository at:

  git://people.freedesktop.org/~gabbayo/linux tags/amdkfd-fixes-2014-12-30

for you to fetch changes up to 38c2adfb00db045a876dd667040abc01b788ad61:

  drm/radeon: Init amdkfd only if it was compiled (2014-12-29 14:46:00 +0200)

----------------------------------------------------------------
Ben Goz (1):
      amdkfd: Fixing topology bug in building sysfs nodes

Oded Gabbay (5):
      amdkfd: Fix accounting of device queues
      amdkfd: Remove duplicate include
      drivers: Move iommu/ before gpu/ in Makefile
      drm: Put amdkfd before radeon in drm Makefile
      drm/radeon: Init amdkfd only if it was compiled

Sasha Levin (1):
      amdkfd: actually allocate longs for the pasid bitmask

 drivers/Makefile                                      |  6 ++++--
 drivers/gpu/drm/Makefile                              |  2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_chardev.c              |  1 -
 drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 13 +++++++++++--
 drivers/gpu/drm/amd/amdkfd/kfd_pasid.c                |  2 +-
 drivers/gpu/drm/amd/amdkfd/kfd_topology.c             |  2 +-
 drivers/gpu/drm/radeon/radeon_kfd.c                   | 12 ++++++++++++
 7 files changed, 30 insertions(+), 8 deletions(-)
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PULL] amdkfd-fixes
  2015-01-06 17:18 ` Oded Gabbay
@ 2015-01-06 17:37   ` Christian König
  0 siblings, 0 replies; 5+ messages in thread
From: Christian König @ 2015-01-06 17:37 UTC (permalink / raw)
  To: Oded Gabbay, David Airlie; +Cc: Alex Deucher, dri-devel

Am 06.01.2015 um 18:18 schrieb Oded Gabbay:
>
>
> On 01/06/2015 07:09 PM, Oded Gabbay wrote:
>> Hi Dave,
>>
>> Here are a couple of bug fixes for the non-HWS path (used for 
>> bring-ups and
>> debugging purposes only).
>>
>> Our QA team run a test on -rc2 and didn't discover any regressions,
>> so I don't expect an additionl pull request for 3.19 unless something
>> unexpected will pop-up.
> Oops, sorry, but I forgot I have an important patch-set regarding 
> fixing the IOCTL error handling as you asked me to do. Christian is 
> still reviewing it but I definitely want it to be in 3.19 as that 
> patch changes the IOCTLs.

Still haven't found time to look closer into it. But feel free to send 
it out cause the general approach sounded good and it is Acked-by: 
Christian König <christian.koenig@amd.com>

Regards,
Christian.

>
>     Oded
>
>>
>>     Oded
>>
>> The following changes since commit 
>> 2f6bd4da08b5054ba933be6f7b17ed02ad6c4162:
>>
>>    Merge tag 'amdkfd-fixes-2014-12-30' of 
>> git://people.freedesktop.org/~gabbayo/linux into linus (2015-01-04 
>> 17:44:43 +1000)
>>
>> are available in the git repository at:
>>
>>
>>    git://people.freedesktop.org/~gabbayo/linux 
>> tags/amdkfd-fixes-2015-01-06
>>
>> for you to fetch changes up to 2030664b709caa769f2b6a1d2e71d8cb343c6884:
>>
>>    drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS) 
>> (2015-01-05 15:48:28 +0200)
>>
>> ----------------------------------------------------------------
>> Alexey Khoroshilov (1):
>>        drm/radeon: do not leave queue acquired if timeout happens in 
>> kgd_hqd_destroy()
>>
>> Ben Goz (4):
>>        drm/amd: Fixing typos in kfd<->kgd interface
>>        drm/amdkfd: Load mqd to hqd in non-HWS mode
>>        drm/radeon: Assign VMID to PASID for IH in non-HWS mode
>>        drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS)
>>
>>   drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 15 
>> +++++++++++++++
>>   drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c          |  2 +-
>>   drivers/gpu/drm/amd/include/kgd_kfd_interface.h       |  2 +-
>>   drivers/gpu/drm/radeon/cikd.h                         |  2 ++
>>   drivers/gpu/drm/radeon/radeon_kfd.c                   | 11 ++++++++---
>>   5 files changed, 27 insertions(+), 5 deletions(-)
>> _______________________________________________
>> dri-devel mailing list
>> dri-devel@lists.freedesktop.org
>> http://lists.freedesktop.org/mailman/listinfo/dri-devel
>>
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PULL] amdkfd-fixes
  2015-01-06 17:09 Oded Gabbay
@ 2015-01-06 17:18 ` Oded Gabbay
  2015-01-06 17:37   ` Christian König
  0 siblings, 1 reply; 5+ messages in thread
From: Oded Gabbay @ 2015-01-06 17:18 UTC (permalink / raw)
  To: David Airlie; +Cc: Alex Deucher, dri-devel



On 01/06/2015 07:09 PM, Oded Gabbay wrote:
> Hi Dave,
>
> Here are a couple of bug fixes for the non-HWS path (used for bring-ups and
> debugging purposes only).
>
> Our QA team run a test on -rc2 and didn't discover any regressions,
> so I don't expect an additionl pull request for 3.19 unless something
> unexpected will pop-up.
Oops, sorry, but I forgot I have an important patch-set regarding fixing the 
IOCTL error handling as you asked me to do. Christian is still reviewing it but 
I definitely want it to be in 3.19 as that patch changes the IOCTLs.

	Oded

>
> 	Oded
>
> The following changes since commit 2f6bd4da08b5054ba933be6f7b17ed02ad6c4162:
>
>    Merge tag 'amdkfd-fixes-2014-12-30' of git://people.freedesktop.org/~gabbayo/linux into linus (2015-01-04 17:44:43 +1000)
>
> are available in the git repository at:
>
>
>    git://people.freedesktop.org/~gabbayo/linux tags/amdkfd-fixes-2015-01-06
>
> for you to fetch changes up to 2030664b709caa769f2b6a1d2e71d8cb343c6884:
>
>    drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS) (2015-01-05 15:48:28 +0200)
>
> ----------------------------------------------------------------
> Alexey Khoroshilov (1):
>        drm/radeon: do not leave queue acquired if timeout happens in kgd_hqd_destroy()
>
> Ben Goz (4):
>        drm/amd: Fixing typos in kfd<->kgd interface
>        drm/amdkfd: Load mqd to hqd in non-HWS mode
>        drm/radeon: Assign VMID to PASID for IH in non-HWS mode
>        drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS)
>
>   drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 15 +++++++++++++++
>   drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c          |  2 +-
>   drivers/gpu/drm/amd/include/kgd_kfd_interface.h       |  2 +-
>   drivers/gpu/drm/radeon/cikd.h                         |  2 ++
>   drivers/gpu/drm/radeon/radeon_kfd.c                   | 11 ++++++++---
>   5 files changed, 27 insertions(+), 5 deletions(-)
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel
>
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PULL] amdkfd-fixes
@ 2015-01-06 17:09 Oded Gabbay
  2015-01-06 17:18 ` Oded Gabbay
  0 siblings, 1 reply; 5+ messages in thread
From: Oded Gabbay @ 2015-01-06 17:09 UTC (permalink / raw)
  To: David Airlie; +Cc: Alex Deucher, dri-devel

Hi Dave,

Here are a couple of bug fixes for the non-HWS path (used for bring-ups and 
debugging purposes only). 

Our QA team run a test on -rc2 and didn't discover any regressions, 
so I don't expect an additionl pull request for 3.19 unless something 
unexpected will pop-up.

	Oded

The following changes since commit 2f6bd4da08b5054ba933be6f7b17ed02ad6c4162:

  Merge tag 'amdkfd-fixes-2014-12-30' of git://people.freedesktop.org/~gabbayo/linux into linus (2015-01-04 17:44:43 +1000)

are available in the git repository at:


  git://people.freedesktop.org/~gabbayo/linux tags/amdkfd-fixes-2015-01-06

for you to fetch changes up to 2030664b709caa769f2b6a1d2e71d8cb343c6884:

  drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS) (2015-01-05 15:48:28 +0200)

----------------------------------------------------------------
Alexey Khoroshilov (1):
      drm/radeon: do not leave queue acquired if timeout happens in kgd_hqd_destroy()

Ben Goz (4):
      drm/amd: Fixing typos in kfd<->kgd interface
      drm/amdkfd: Load mqd to hqd in non-HWS mode
      drm/radeon: Assign VMID to PASID for IH in non-HWS mode
      drm/amdkfd: unmap VMID<-->PASID when relesing VMID (non-HWS)

 drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 15 +++++++++++++++
 drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c          |  2 +-
 drivers/gpu/drm/amd/include/kgd_kfd_interface.h       |  2 +-
 drivers/gpu/drm/radeon/cikd.h                         |  2 ++
 drivers/gpu/drm/radeon/radeon_kfd.c                   | 11 ++++++++---
 5 files changed, 27 insertions(+), 5 deletions(-)
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PULL] amdkfd-fixes
@ 2014-12-23 21:40 Oded Gabbay
  0 siblings, 0 replies; 5+ messages in thread
From: Oded Gabbay @ 2014-12-23 21:40 UTC (permalink / raw)
  To: airlied, dri-devel; +Cc: alexander.deucher, Dana.Elifaz

Hi Dave,

Two fixes for amdkfd.

- Display MEC fw version in topology. Without this, the HSA userspace
  stack is broken.

- Init apertures information only once per process

The following changes since commit 97bf6af1f928216fd6c5a66e8a57bfa95a659672:

  Linux 3.19-rc1 (2014-12-20 17:08:50 -0800)

are available in the git repository at:

  git://people.freedesktop.org/~gabbayo/linux tags/amdkfd-fixes-2014-12-23

for you to fetch changes up to dd59239a9862a42e4b8d47e4aaa8d595d08c29ab:

  amdkfd: init aperture once per process (2014-11-18 13:56:23 +0200)

----------------------------------------------------------------
Alexey Skidanov (1):
      amdkfd: init aperture once per process

Oded Gabbay (3):
      drm/amd: Add get_fw_version to kfd-->kgd interface
      drm/radeon: Add implementation of get_fw_version
      amdkfd: Display MEC fw version in topology node

 drivers/gpu/drm/amd/amdkfd/kfd_chardev.c        |  4 --
 drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c    |  6 +--
 drivers/gpu/drm/amd/amdkfd/kfd_process.c        |  9 +++++
 drivers/gpu/drm/amd/amdkfd/kfd_topology.c       |  8 +++-
 drivers/gpu/drm/amd/include/kgd_kfd_interface.h | 15 +++++++
 drivers/gpu/drm/radeon/radeon_kfd.c             | 53 +++++++++++++++++++++++++
 6 files changed, 85 insertions(+), 10 deletions(-)
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2015-01-06 17:38 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-12-30 19:35 [PULL] amdkfd-fixes Oded Gabbay
  -- strict thread matches above, loose matches on Subject: below --
2015-01-06 17:09 Oded Gabbay
2015-01-06 17:18 ` Oded Gabbay
2015-01-06 17:37   ` Christian König
2014-12-23 21:40 Oded Gabbay

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.