- 06 Dec, 2017 40 commits
-
-
Christian König authored
Instead of the global statistics use the per context bytes moved counter. v2: rebased Signed-off-by:
Christian König <christian.koenig@amd.com> Reviewed-by:
Michel Dänzer <michel.daenzer@amd.com> Reviewed-by:
Chunming Zhou <david1.zhou@amd.com> Tested-by:
Dieter Nützel <Dieter@nuetzel-hh.de> Tested-by:
Felix Kuehling <Felix.Kuehling@amd.com> Signed-off-by:
Alex Deucher <alexander.deucher@amd.com>
-
Christian König authored
This way we can finally use some more stats. Signed-off-by:
-
Christian König authored
Fixes interface changes done in the following commits: drm/ttm: add operation ctx to ttm_bo_validate v2 drm/ttm: add context to driver move callback as well I missed this driver because it is in the staging area. Acked-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Acked-by:
Hans de Goede <hdegoede@redhat.com> Signed-off-by:
-
Christian König authored
Add some statistics how many bytes we have moved. Signed-off-by:
-
Christian König authored
Instead of passing the parameters manually. Signed-off-by:
-
Christian König authored
Instead of passing down the parameters manually to every function. Signed-off-by:
-
Christian König authored
Instead of specifying interruptible and no_wait_gpu manually. v2: rebase Signed-off-by:
-
Christian König authored
Instead of specifying if sleeping should be interruptible. Signed-off-by:
-
Christian König authored
Give moving a BO into place an operation context to work with. v2: rebased Signed-off-by:
-
Roger He authored
to indicate page order for each element in the pool Reviewed-by:
Roger He <Hongbo.He@amd.com> Signed-off-by:
-
Roger He authored
Reviewed-by:
-
Ernst Sjöstrand authored
This is a followup to: drm/amd/powerplay: Fix buffer overflows with mc_reg_address Rework *_set_mc_special_registers for the other architectures to use the same logic as the first patch. This allows the last entry of the array to be filled without an error message for example. This doesn't fix any known problems, perhaps avoided by luck. Reviewed-by:
Ernst Sjöstrand <ernstp@gmail.com> Signed-off-by:
-
Ernst Sjöstrand authored
Smatch warned about the following lines: ci_set_mc_special_registers() error: buffer overflow 'table->mc_reg_address' 16 <= 16 tonga_set_mc_special_registers() error: buffer overflow 'table->mc_reg_address' 16 <= 16 Change the logic to check before access instead of after incrementing. It's fine if j reaches max after we're done. This allows the last entry of the array to be filled without an error message for example. Changed some whitespace to clarify grouping. Reviewed-by:
-
Ernst Sjöstrand authored
Reported by smatch: amdgpu_atombios_i2c_process_i2c_ch() error: we previously assumed 'buf' could be null Reviewed-by:
-
Ernst Sjöstrand authored
Reviewed-by:
-
Ernst Sjöstrand authored
Reported by smatch: init_overdrive_limits() error: uninitialized symbol 'result'. get_clock_voltage_dependency_table() warn: inconsistent indenting v2: set result to 0 (Alex) Reviewed-by:
-
Christian König authored
Turned out that VCE still has a placement restriction that BOs can't cross a 4GB boundary. Fix this by adding a command submission parser prepass to correctly place the buffers. v2: add function description Signed-off-by:
-
Christian König authored
For VCE to work properly the start of the GTT space must be aligned to a 4GB boundary. v2: add comment why we do this Signed-off-by:
-
Christian König authored
Remove some outdated comments and all code which tries to reduce the VRAM size mapped into the MC. This is superfluous and misleading since we never actually program the size. v2: handle gmc_v6_0.c as well Signed-off-by:
-
Joe Perches authored
Line continuations with excess spacing causes unexpected output. Miscellanea: o Added missing '\n' to a few of the coalesced pr_<level> formats Reviewed-by:
Harry Wentland <harry.wentland@amd.com> Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Gustavo A. R. Silva authored
Make use of the swap macro instead of _manually_ swapping values and remove unnecessary variable swap. This makes the code easier to read and maintain. This code was detected with the help of Coccinelle. Signed-off-by:
Gustavo A. R. Silva <garsilva@embeddedor.com> Signed-off-by:
-
Gustavo A. R. Silva authored
Make use of the swap macro instead of _manually_ swapping values and remove unnecessary variable temp. This makes the code easier to read and maintain. This code was detected with the help of Coccinelle. Signed-off-by:
-
Alex Deucher authored
To avoid spamming the logs on non-ECC boards. Reviewed-by:
-
Christian König authored
Fixing warning/compile errors on 32bit kernels. Signed-off-by:
-
Colin Ian King authored
Use kcalloc instead of kzalloc and the cast on the return from kzalloc is unnecessary and can be removed. Reviewed-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Christian König authored
Not used any more. Signed-off-by:
-
Christian König authored
Extern is the default for function declerations anyway. Signed-off-by:
-
Christian König authored
Extern is the default for function declerations anyway and this solves a bunch of 80char per line issues. Signed-off-by:
-
Christian König authored
There is no guarantee that the next entry on the ddelete list stays on the list when we drop the locks. Completely rework this mess by moving processed entries on a temporary list. Signed-off-by:
-
Christian König authored
The code path isn't hit at the moment, but we need to take the lock to add the BO back to the LRU. Signed-off-by:
-
Christian König authored
Don't even try to resize the BAR when there is no window above 4GB. Signed-off-by:
-
Shirish S authored
This patch fixes static checker warning of "warn: cast after binop" introduced by 56087b31 drm/amd/display: fix high part address in dm_plane_helper_prepare_fb() Signed-off-by:
Shirish S <shirish.s@amd.com> Reviewed-by:
-
Monk Liu authored
virtual DCE Timer structure is already released after its sw_fini(), so we need to cancel the its Timer in hw_fini() otherwise the Timer canceling is missed. v2: use for loop and num_crtc to replace original code Signed-off-by:
Monk Liu <Monk.Liu@amd.com> Reviewed-by:
-
Monk Liu authored
this fix the issue that access memory after freed after driver unloaded. Signed-off-by:
-
Monk Liu authored
Signed-off-by:
-
Monk Liu authored
use bo_create/free_kernel instead of manually doing it Signed-off-by:
-
Monk Liu authored
instead of doing it in each GFX ip's sw_fini Signed-off-by:
-
Monk Liu authored
Signed-off-by:
-
Monk Liu authored
NULL pointer is because original logic will step into set_pde_pte() even after the gart.ptr is freed due to there are twice gart_unbind() on all gart area. also, there are other minor fixes: 1,since gart_init only create dummy page, the corresponding gart_fini shouldn't do more like unbinding all GART, this is unnecessary because in driver fini stage all GART unbinding had already been done during each IP's SW_FINI (GMC's SW_FINI is the last one called), so remove the step for the GART unbinding in gart_fini(). 2,gart_fini() is already invoked during each GMC IP's gart_fini routine,e.g. gmc_vx_0_gart_fini(), so no need to manually call it during ttm_fini(). 3,amdgpu_gem_force_release() should be put ahead of amdgpu_vm_manager_fini() Signed-off-by:
-
Monk Liu authored
for gart_ram_alloc/free, they are never used in driver thus ripe them out totally. for gart_vram_pin/unpin, they are not needed becuase we can use bo_creat_kernel/free to replace the original manual way in the gart_vram_alloc/free, thus gart_vram_pin/unpin can also be riped out. Signed-off-by:
-
