In Vulkan we have extensions to assist with capture in replay in a
world where addresses are returned to the application. This involves
creating buffers at the same VA during replay as they were during
capture.
By itself libdrm_amdgpu already has support for this, but there is
the obvious failure mode that if another buffer is already allocated
at that VA things fail spectacularly. This is an actual issue as
internal buffers, like winsys images or shader binaries also
participate in the same VA allocation.
To avoid this problem applications have to create buffers which
are going to be captured with a flag, and the implementation is to
separate VA allocation for those buffers to reduce the collision risk:
"Implementations are expected to separate such buffers in the GPU address
space so normal allocations will avoid using these addresses. Apps/tools
should avoid mixing app-provided and implementation-provided addresses for
buffers created with VK_BUFFER_CREATE_DEVICE_ADDRESS_CAPTURE_REPLAY_BIT,
to avoid address space allocation conflicts."
This patch implements that by adding a flag for these buffers and allocating
address space from the top of the address range instead of the bottom.
Signed-off-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Christian König <christian.koenig@amd.com>
This allows users to select the library type (static or shared)
using the Meson -Ddefault_library built-in option.
Issue: https://gitlab.freedesktop.org/mesa/drm/-/issues/45
Reviewed-by: Simon Ser <contact@emersion.fr>
Signed-off-by: Fang Tan <tanfang@uniontech.com>
The amdgpu implementation uses `clock_gettime` so it needs to check whether it needs to link
against `-librt`.
Signed-off-by: Valentin Churavy <v.churavy@gmail.com>
Mesa expects to be able to map the same buffer, without unmapping it.
This leads to problem on long-running program.
On the other hand, libdrm uses cpu_map_count as a refcount and expects
its value to decrease so it can unmap buffers.
The previoulsy proprosed fix (https://patchwork.freedesktop.org/patch/258005/)
stopped increased the counter when it went past INT_MAX.
This commit instead proposes to use a larger type to store cpu_map_count.
The outcome is the same: long running apps will not crash, only the
implementation differs.
Closes: https://gitlab.freedesktop.org/mesa/mesa/issues/1423
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
This patch is to use generic variables as the input of amdgpu_cs_submit_raw2.
Because amdgpu_cs_submit_one won't handle IOCTL directly.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
So far, amdgpu_cs_submit_raw2 is mainly used for upper layer (Mesa), however,
amdgpu_cs_submit is used for current all unit tests. Our intention is that the
unit tests can actually verify the API which is really used.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Use alloca instead of malloc, then we don't need free them at the end of this
function.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Combined with -Wundef (added in 75758d2ccf & enforced in ba17673eed),
this provides absolute safety against #ifdef typos.
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
v2: nit-picks fix
Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Christian König <Christian.Koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
For the xf86drm.[ch] part : Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
None of the tests are bash specific. Tested with bash, zsh, dash, mksh
and ksh.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Tested-by: Niclas Zeising <zeising@daemonic.se>
Reviewed-by: Niclas Zeising <zeising@daemonic.se>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Simplifies its callers.
dev->bo_table_mutex is now always held when amdgpu_bo_create is called
(this was already the case in amdgpu_bo_import).
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
And propagate drmIoctl's return value.
This allows replacing all remaining open-coded DRM_IOCTL_GEM_CLOSE
ioctl calls with amdgpu_close_kms_handle calls.
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Avoids compiler warning:
../../amdgpu/amdgpu_cs.c: In function 'amdgpu_cs_ctx_override_priority':
../../amdgpu/amdgpu_cs.c:155:2: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
union drm_amdgpu_sched args;
^~~~~
Reviewed-by: Christian König <christian.koenig@amd.com>
v2: adapt to new one transfer ioctl
Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
v2: symbos are stored in lexical order.
v3: drop export/import and extra query indirection
Signed-off-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Clang complains when initializing unions using "= {0}"
so instead use memset.
Cc: Sean Paul <seanpaul@chromium.org>
Cc: Alistair Strachan <astrachan@google.com>
Cc: Marissa Wall <marissaw@google.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: John Stultz <john.stultz@linaro.org>
This way we can override the priority of a single context using a
master fd.
Since we cannot usefully create an amdgpu device of a master fd
without the fd deduplication kicking in this takes a plain fd.
This can be used by e.g. radv to get high priority contexts using
a master fd from the primary node or a lease.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Saw a couple of typos fixes in the patch DragonFlyBSD carries [1], so
I ran codespell (a spell checker for code) on the whole repo.
[1] https://github.com/DragonFlyBSD/DPorts/blob/master/graphics/libdrm/files/patch-xf86drm.c
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
It will make bugs like the one fixed with previous patch dead obvious.
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Set/export the NM variable since it may not be set already.
Fixes: 4f08bfe96d ("*-symbol-check: Don't hard-code nm executable")
Cc: Heiko Becker <heirecka@exherbo.org>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Now that symbols that should be exported are annotated accordingly, make
all the rest hidden by default.
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Now that symbols that should be exported are annotated accordingly, make
all the rest hidden by default.
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
This was done with:
nm --dynamic --defined-only build/amdgpu/libdrm_amdgpu.so | \
grep amdgpu_ | \
cut -d' ' -f3 > /tmp/a.txt
while read sym; do
read f func line _ <<<$(cscope -d -L -1 $sym)
if [ ! -z "$f" ]; then
line=$((line-1))
sed -i "${line}s/^/drm_public /" $f
fi
done < /tmp/a.txt
Then the alignment of function arguments were manually fixed all over.
The idea here will be to switch the default visibility to hidden so we
don't export symbols we shouldn't.
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
The heap is checked by the kernel and not libdrm, to make it even worse
it prevented allocating resources other than VRAM and GTT.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
For Pro OGL be able to work with upstream libdrm.
Signed-off-by: Qiang Yu <Qiang.Yu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
If nothing is found, error should be returned.
v2: udpate the error value different from parameter check
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
a helper function to create and initialize amdgpu bo
v2: update error handling: add label and free bo
v3: update error handling: separate each error label
v4: update error handling and rebase
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Fix potential memory leak when handle flink bo in bo import.
Free the flink bo after bo import and in error handling.
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Arithmetic using void* pointers isn't defined by the C standard, only as
a GCC extension. Avoids compiler warnings:
../../amdgpu/amdgpu_bo.c: In function ‘amdgpu_find_bo_by_cpu_mapping’:
../../amdgpu/amdgpu_bo.c:554:48: warning: pointer of type ‘void *’ used in arithmetic [-Wpointer-arith]
if (cpu >= bo->cpu_ptr && cpu < (bo->cpu_ptr + bo->alloc_size))
^
../../amdgpu/amdgpu_bo.c:561:23: warning: pointer of type ‘void *’ used in subtraction [-Wpointer-arith]
*offset_in_bo = cpu - bo->cpu_ptr;
^
v2: Use uintptr_t instead of char*, don't change function signature
(Junwei Zhang)
Fixes: 4d454424e1 ("amdgpu: add a function to find bo by cpu mapping
(v2)")
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Stupid me, max_key must always be larger than key.
Signed-off-by: Christian König <christian.koenig@amd.com>
Bugzilla: https://bugs.freedesktop.org/107552
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
The compiler points out that an int doesn't work as intended if
dev->bo_handles.max_key > INT_MAX:
../../amdgpu/amdgpu_bo.c: In function ‘amdgpu_find_bo_by_cpu_mapping’:
../../amdgpu/amdgpu_bo.c:550:16: warning: comparison of integer expressions of different signedness: ‘int’ and ‘uint32_t’ {aka ‘unsigned int’} [-Wsign-compare]
for (i = 0; i < dev->bo_handles.max_key; i++) {
^
../../amdgpu/amdgpu_bo.c:558:8: warning: comparison of integer expressions of different signedness: ‘int’ and ‘uint32_t’ {aka ‘unsigned int’} [-Wsign-compare]
if (i < dev->bo_handles.max_key) {
^
Fixes: 4d454424e1 ("amdgpu: add a function to find bo by cpu mapping
(v2)")
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Userspace needs to know if the user memory is from BO or malloc.
v2: update mutex range and rebase
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
When create bo from user memory, add it to handle table
for future query.
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Use the correct files to build libdrm_amdgpu.
Signed-of-by: Mike Lothian <mike@fireburn.co.uk>
Fixes: d6cb0ee408 ("amdgpu: remove the hash table implementation")
Cc: Christian König <christian.koenig@amd.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
This way we can always find a BO structure by its handle.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-and-Tested-by: Junwei Zhang <Jerry.Zhang@amd.com>
Bas Nieuwenhuizen
Leo Liu
Fang Tan
Valentin Churavy
Pierre-Eric Pelloux-Prayer
Huang Rui
Eric Engestrom
Chunming Zhou
Marek Olšák
Emil Velikov
Michel Dänzer
John Stultz
Emil Velikov
Lucas De Marchi
Christian König
Qiang Yu
Junwei Zhang
Mike Lothian