Re: [PATCH 03/14] drm/amdkfd: map multiple processes to HW scheduler

2017-12-05 Thread Oded Gabbay
On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling wrote: > Allow HWS to to execute multiple processes on the hardware > concurrently. The number of concurrent processes is limited by > the number of VMIDs allocated to the HWS. > > A module parameter can be used for limiting this further or turn > it

Re: [PATCH 04/14] drm/amdkfd: Fix oversubscription accounting

2017-12-05 Thread Oded Gabbay
On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling wrote: > Don't count SDMA queues towards compute HQD oversubscription when > deciding whether to create a chained runlist. > > Signed-off-by: Jay Cornwall > Signed-off-by: Felix Kuehling > --- > drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c | 5

Re: [PATCH 05/14] drm/amdgpu: Fix definition of KFD_CIK_SDMA_QUEUE_OFFSET

2017-12-05 Thread Oded Gabbay
On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling wrote: > This counts the queue offset in register index, not register address. > > Signed-off-by: Felix Kuehling > --- > drivers/gpu/drm/amd/amdgpu/cikd.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/gpu/drm/amd/

Re: [PATCH 06/14] drm/amdgpu: Add kfd2kgd APIs for dumping HQDs

2017-12-05 Thread Oded Gabbay
On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling wrote: > This can be used by KFD for debugging features, such as dumping > HQDs in debugfs. > > Signed-off-by: Felix Kuehling > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c | 71 > drivers/gpu/drm/amd/amdgpu/amdgpu_a

Re: [PATCH 07/14] drm/amdkfd: Add debugfs support to KFD

2017-12-05 Thread Oded Gabbay
On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling wrote: > This commit adds several debugfs entries for kfd: > > kfd/hqds: dumps all HQDs on all GPUs for KFD-controlled compute and > SDMA RLC queues > > kfd/mqds: dumps all MQDs of all KFD processes on all GPUs > > kfd/rls: dumps HWS runlists on

Re: [PATCH 08/14] drm/amdkfd: Get reference to lead_thread task struct

2017-12-05 Thread Oded Gabbay
On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling wrote: > Increment the kfd_process.lead_thread's reference counter to make > it safe to dereference. This is needed for getting a safe reference > to the process' mm_struct. I don't object to this patch, but I thought we don't dereference the proces

Re: [PATCH 0/2] Move scheduler out of AMDGPU

2017-12-05 Thread Lucas Stach
Hi Alex, Am Montag, den 04.12.2017, 16:47 -0500 schrieb Alex Deucher: > On Fri, Dec 1, 2017 at 10:55 AM, Christian König > > wrote: > > Am 01.12.2017 um 16:28 schrieb Lucas Stach: > > > > > > Hi all, > > > > > > so this is the first step to make the marvelous AMDGPU scheduler useable > > > for

[PATCH] drm/ttm: swap consecutive allocated pooled pages v4

2017-12-05 Thread Christian König
When we detect consecutive allocation of pages swap them to avoid accidentally freeing them as huge page. v2: use swap v3: check if it's really the first allocated page v4: don't touch the loop variable Signed-off-by: Christian König --- drivers/gpu/drm/ttm/ttm_page_alloc.c | 11 +-- 1

Deadlocks with multiple applications on AMD RX460

2017-12-05 Thread Luís Mendes
Hi, I don't know how to start with, there are multiple issues, but maybe I can start with the issue in amdgpu driver in linux-4.15-rc2. There is some patch that is quite relevant for the AMD RX460 to initialize properly which is in both the development branches ~agd5f/amd-staging-drm-next (which

[PATCH libdrm] amdgpu.ids: Refresh from AMD 17.40 release

2017-12-05 Thread Michel Dänzer
From: Michel Dänzer Reviewed-by: Alex Deucher Signed-off-by: Michel Dänzer --- data/amdgpu.ids | 31 +-- 1 file changed, 29 insertions(+), 2 deletions(-) diff --git a/data/amdgpu.ids b/data/amdgpu.ids index f6c65dd9..1828e410 100644 --- a/data/amdgpu.ids +++ b/data

Re: [PATCH libdrm] amdgpu.ids: Refresh from AMD 17.40 release

2017-12-05 Thread Christian König
Am 05.12.2017 um 12:59 schrieb Michel Dänzer: From: Michel Dänzer Reviewed-by: Alex Deucher Signed-off-by: Michel Dänzer Acked-by: Christian König --- data/amdgpu.ids | 31 +-- 1 file changed, 29 insertions(+), 2 deletions(-) diff --git a/data/amdgpu.ids b

[PATCH] drm/amdgpu: Fix amdgpu_sync_add_later to preserve explicit flag.

2017-12-05 Thread Andrey Grodzovsky
Signed-off-by: Andrey Grodzovsky --- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c | 9 ++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index f3d1a25..84fd367 100644 --- a/drivers/gpu/drm/amd/a

Re: [PATCH] drm/amdgpu: Fix amdgpu_sync_add_later to preserve explicit flag.

2017-12-05 Thread Christian König
Am 05.12.2017 um 15:41 schrieb Andrey Grodzovsky: Signed-off-by: Andrey Grodzovsky Reviewed-by: Christian König --- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c | 9 ++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gp

Re: [PATCH] drm/ttm: swap consecutive allocated pooled pages v4

2017-12-05 Thread Michel Dänzer
On 2017-12-05 01:54 PM, Christian König wrote: > When we detect consecutive allocation of pages swap them to avoid > accidentally freeing them as huge page. > > v2: use swap > v3: check if it's really the first allocated page > v4: don't touch the loop variable > > Signed-off-by: Christian König

Deadlocks with multiple applications on AMD RX460

2017-12-05 Thread Luís Mendes
Hi, I don't know how to start with, there are multiple issues, but maybe I can start with the issue in amdgpu driver in linux-4.15-rc2. There is some patch that is quite relevant for the AMD RX460 to initialize properly which is in both the development branches ~agd5f/amd-staging-drm-next (which

Re: [PATCH] drm: amd: dc: don't use FP math when Kcov is enabled

2017-12-05 Thread Arnd Bergmann
On Mon, Dec 4, 2017 at 9:34 PM, Harry Wentland wrote: > On 2017-12-04 08:08 AM, Arnd Bergmann wrote: >> Building the DCN 1.0 Raven display driver with CONFIG_KCOV_INSTRUMENT_ALL=y >> and CONFIG_KCOV_ENABLE_COMPARISONS=y results in warnings about many functions >> that do a comparison of floating-p

Re: [PATCH] drm: amd: dc: don't use FP math when Kcov is enabled

2017-12-05 Thread Harry Wentland
On 2017-12-05 07:05 AM, Arnd Bergmann wrote: > On Mon, Dec 4, 2017 at 9:34 PM, Harry Wentland wrote: >> On 2017-12-04 08:08 AM, Arnd Bergmann wrote: >>> Building the DCN 1.0 Raven display driver with CONFIG_KCOV_INSTRUMENT_ALL=y >>> and CONFIG_KCOV_ENABLE_COMPARISONS=y results in warnings about ma

[PATCH umr] Add initial SDMA ring decoding support

2017-12-05 Thread Tom St Denis
Tested on a Polaris10 running testdma with mesa. Based on packet descriptions found in tonga_sdma_pkt_open.h found in the kernel tree. Like the other ring decoder also supports: follow_ib, bits, and use_colour. Signed-off-by: Tom St Denis --- src/app/ring_read.c | 5 + src/lib/dump_ib.c

Re: [PATCH umr] Add initial SDMA ring decoding support

2017-12-05 Thread Alex Deucher
On Tue, Dec 5, 2017 at 12:03 PM, Tom St Denis wrote: > Tested on a Polaris10 running testdma with mesa. Based on packet > descriptions found in tonga_sdma_pkt_open.h found in the kernel tree. > > Like the other ring decoder also supports: follow_ib, bits, and use_colour. > > Signed-off-by: Tom S

Re: [PATCH 08/14] drm/amdkfd: Get reference to lead_thread task struct

2017-12-05 Thread Felix Kuehling
On 2017-12-05 03:43 AM, Oded Gabbay wrote: > On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling > wrote: >> Increment the kfd_process.lead_thread's reference counter to make >> it safe to dereference. This is needed for getting a safe reference >> to the process' mm_struct. > I don't object to this

Re: [PATCH 04/14] drm/amdkfd: Fix oversubscription accounting

2017-12-05 Thread Felix Kuehling
On 2017-12-05 03:10 AM, Oded Gabbay wrote: > On Tue, Nov 28, 2017 at 1:29 AM, Felix Kuehling > wrote: >> Don't count SDMA queues towards compute HQD oversubscription when >> deciding whether to create a chained runlist. >> >> Signed-off-by: Jay Cornwall >> Signed-off-by: Felix Kuehling >> --- >

Re: [PATCH 0/2] Move scheduler out of AMDGPU

2017-12-05 Thread Alex Deucher
On Tue, Dec 5, 2017 at 4:18 AM, Lucas Stach wrote: > Hi Alex, > > Am Montag, den 04.12.2017, 16:47 -0500 schrieb Alex Deucher: >> On Fri, Dec 1, 2017 at 10:55 AM, Christian König >> > wrote: >> > Am 01.12.2017 um 16:28 schrieb Lucas Stach: >> > > >> > > Hi all, >> > > >> > > so this is the first

Re: [PATCH] drm/amdgpu: Fix amdgpu_sync_add_later to preserve explicit flag.

2017-12-05 Thread William Lewis
On 12/05/2017 08:41 AM, Andrey Grodzovsky wrote: > Signed-off-by: Andrey Grodzovsky > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c | 9 ++--- > 1 file changed, 6 insertions(+), 3 deletions(-) > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c > b/drivers/gpu/drm/amd/amdgpu/amdgp

RE: [PATCH] drm/ttm: swap consecutive allocated pooled pages v4

2017-12-05 Thread He, Roger
Reviewed-by: Roger He Thanks Roger(Hongbo.He) -Original Message- From: Christian König [mailto:ckoenig.leichtzumer...@gmail.com] Sent: Tuesday, December 05, 2017 8:55 PM To: dri-de...@lists.freedesktop.org; amd-gfx@lists.freedesktop.org; He, Roger Subject: [PATCH] drm/ttm: swap consecu