SWDEV-133886 - [Linux-Pro][VEGA ][Nuke]OpenCL kernel compilation failed" error is observed with Nuke on Vega10
- Port CL#1116855 to OCL/ROCm.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocprogram.cpp#73 edit
Use mbind to specify the NUMA node for system memory allocation. This
only works with HSA_USERPTR_FOR_PAGED_MEM=1.
Change-Id: I88e7815d5a5aefcc4c22358c1a4a1635d7677ef3
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Use mbind to specify the NUMA node for system memory allocation. This
only works with HSA_USERPTR_FOR_PAGED_MEM=1.
Change-Id: I88e7815d5a5aefcc4c22358c1a4a1635d7677ef3
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
[ROCm/ROCR-Runtime commit: cb4814eadc]
Marking heap memory as executable using mprotect() is not allowed
by SELinux. mprotect() calls that try to do this will fail on systems
with SELinux enabled. This is also a security risk, so it should be
fixed even on systems that allow this.
Any memory we want to mark as executable must be allocated using mmap().
See https://www.akkadia.org/drepper/selinux-mem.html
The two places where we try to mark heap memory as executable both use
posix_memalign() to allocate the heap memory. In both cases, the
alignment value passed into this function is always equal to PAGE_SIZE,
which means that they are safe to replace with mmap(), which guarantees
alignment to PAGE_SIZE. In this case PAGE_SIZE has been set to
sysconf(_SC_PAGESIZE);
v2:
- Use MAP_PRIVATE instead of MAP_SHARED. This matches the behavior
of memory allocated by posix_memalign()
- Ignore alignment hints instead of returning error when we can't
accommodate them.
- Drop alignment parameter of allocate_exec_aligned_memory() since
the only alignment supported is sysconf(_SC_PAGESIZE).
- Remove extra parameter from fmm_release().
- Add error path to fmm_allocate_host_cpu() for when mmap fails.
v3:
- Avoid use after free.
Change-Id: I7d51279790d9700bc3fa761c44bfde1c1936019b
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Marking heap memory as executable using mprotect() is not allowed
by SELinux. mprotect() calls that try to do this will fail on systems
with SELinux enabled. This is also a security risk, so it should be
fixed even on systems that allow this.
Any memory we want to mark as executable must be allocated using mmap().
See https://www.akkadia.org/drepper/selinux-mem.html
The two places where we try to mark heap memory as executable both use
posix_memalign() to allocate the heap memory. In both cases, the
alignment value passed into this function is always equal to PAGE_SIZE,
which means that they are safe to replace with mmap(), which guarantees
alignment to PAGE_SIZE. In this case PAGE_SIZE has been set to
sysconf(_SC_PAGESIZE);
v2:
- Use MAP_PRIVATE instead of MAP_SHARED. This matches the behavior
of memory allocated by posix_memalign()
- Ignore alignment hints instead of returning error when we can't
accommodate them.
- Drop alignment parameter of allocate_exec_aligned_memory() since
the only alignment supported is sysconf(_SC_PAGESIZE).
- Remove extra parameter from fmm_release().
- Add error path to fmm_allocate_host_cpu() for when mmap fails.
v3:
- Avoid use after free.
Change-Id: I7d51279790d9700bc3fa761c44bfde1c1936019b
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
[ROCm/ROCR-Runtime commit: e2ed9cf79a]
src/perfctr.c: In function ‘destroy_shared_region’:
src/perfctr.c:154:10: error: logical ‘and’ of equal expressions [-Werror=logical-op]
if (sem && sem != SEM_FAILED) {
^~
src/perfctr.c: In function ‘update_block_slots’:
src/perfctr.c:323:11: error: logical ‘or’ of equal expressions [-Werror=logical-op]
if (!sem || sem == SEM_FAILED)
^~
v2:
- Initialize and reset sem to SEM_FAILED.
Change-Id: Id70361079b715c4946b13e4460e4fd85d9542c46
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
src/perfctr.c: In function ‘destroy_shared_region’:
src/perfctr.c:154:10: error: logical ‘and’ of equal expressions [-Werror=logical-op]
if (sem && sem != SEM_FAILED) {
^~
src/perfctr.c: In function ‘update_block_slots’:
src/perfctr.c:323:11: error: logical ‘or’ of equal expressions [-Werror=logical-op]
if (!sem || sem == SEM_FAILED)
^~
v2:
- Initialize and reset sem to SEM_FAILED.
Change-Id: Id70361079b715c4946b13e4460e4fd85d9542c46
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
[ROCm/ROCR-Runtime commit: 4815d9b887]
SWDEV-2 - Change OpenCL version number from 2517 to 2518.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2265 edit
[ROCm/clr commit: c6ab68be17]
SWDEV-2 - Change OpenCL version number from 2517 to 2518.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2265 edit
TensorFlow was running out of VRAM due to padding up allocations
from legacy memory APIs. These allocations have been added to
the fragment allocator to improve VRAM utilization.
Change-Id: Ic680fff576a0434b3b17a4c91746da44e09957fa
TensorFlow was running out of VRAM due to padding up allocations
from legacy memory APIs. These allocations have been added to
the fragment allocator to improve VRAM utilization.
Change-Id: Ic680fff576a0434b3b17a4c91746da44e09957fa
[ROCm/ROCR-Runtime commit: 4f299a9909]
SWDEV-120036 - Supporting the cl_amd_device_attribute_query on the ROC device
- divided the query of HSA_AMD_AGENT_INFO_MEMORY_WIDTH by 32 to obtained the globale memory channels as done in ORCA and PAL
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#69 edit
[ROCm/clr commit: 05216e1f54]
SWDEV-120036 - Supporting the cl_amd_device_attribute_query on the ROC device
- divided the query of HSA_AMD_AGENT_INFO_MEMORY_WIDTH by 32 to obtained the globale memory channels as done in ORCA and PAL
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#69 edit
Fix a while loop that can cause forever loop when cpuid instruction
doesn't work properly.
Change-Id: Iefa49d23b40c994eb4369621974a7d3c4067e47a
Signed-off-by: Amber Lin <Amber.Lin@amd.com>
Fix a while loop that can cause forever loop when cpuid instruction
doesn't work properly.
Change-Id: Iefa49d23b40c994eb4369621974a7d3c4067e47a
Signed-off-by: Amber Lin <Amber.Lin@amd.com>
[ROCm/ROCR-Runtime commit: 5815d9de9b]
[Synopsis]
If any of kernel arguments is a MACRO its location calculation is wrong (location of its definition is actually calculated).
Thus garbage code is being produced on the place of such a MACRO starting from the end of its actual definition.
[Solution]
Add isMacroBodyExpansion and isMacroArgExpansion checks on kernel arguments.
[ROCm/clr commit: f3ef942407]
[Synopsis]
If any of kernel arguments is a MACRO its location calculation is wrong (location of its definition is actually calculated).
Thus garbage code is being produced on the place of such a MACRO starting from the end of its actual definition.
[Solution]
Add isMacroBodyExpansion and isMacroArgExpansion checks on kernel arguments.
[Synopsis]
If any of kernel arguments is a MACRO its location calculation is wrong (location of its definition is actually calculated).
Thus garbage code is being produced on the place of such a MACRO starting from the end of its actual definition.
[Solution]
Add isMacroBodyExpansion and isMacroArgExpansion checks on kernel arguments.
SWDEV-2 - Change OpenCL version number from 2516 to 2517.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2264 edit
[ROCm/clr commit: 469491fd51]
SWDEV-2 - Change OpenCL version number from 2516 to 2517.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2264 edit
SWDEV-2 - Change OpenCL version number from 2515 to 2516.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2263 edit
[ROCm/clr commit: bf1ff9f160]
SWDEV-2 - Change OpenCL version number from 2515 to 2516.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2263 edit
SWDEV-133815 - PAL support for Linux Pro w/OpenCL 2.0 support
- Fix the clear size in FillBuffer. It's in bytes.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#61 edit
[ROCm/clr commit: 7b524705c1]
SWDEV-133815 - PAL support for Linux Pro w/OpenCL 2.0 support
- Fix the clear size in FillBuffer. It's in bytes.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#61 edit
SWDEV-2 - Change OpenCL version number from 2514 to 2515.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2262 edit
[ROCm/clr commit: 3bd2d13b40]
SWDEV-2 - Change OpenCL version number from 2514 to 2515.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2262 edit
foreman
Felix Kuehling
Chris Freehill
Tom Stellard
Sean Keely
Amber Lin
Evgeny Mankov
Ramesh Errabolu
Evgeny
Alex Voicu