Currently COMGR doesn't provide global variable size and runtime
parses ELF binary directly. Avoid parsing for HIP. That can save
5% in hipModuleLoad() time.
Change-Id: I47540d1e957bdb0c2406b6b848222de2920b2504
When the HIP compiler set to a symblink that was created outside
/opt/rocm/llvm/bin, HIP's cmake would get HIP_CLANG_ROOT incorrectly.
This teaches hip-config to get clang's real path first and use that to determine LLVM's directory.
Change-Id: I6275b67f843aff256dc72cf478e575224e0cedc3
During RPM upgrade, lib/libOpenCL.so* files/symlinks from the previous
installed version are not removed.
This change will remove those symlinks as part of the RPM remove
or upgrade operation.
Signed-off-by: Saravanan Solaiyappan <saravanan.solaiyappan@amd.com>
Change-Id: Ia943407052ec24911756042a5f9839a605522443
Pass active queue for transfers in the cache coherency layer.
That will allow to use device transfer queue only for
cases when active queue isn't available, because using device
transfer queue from another active queue may cause a deadlock
Change-Id: Ifbe7e0303b77dbf6eeda3939ffbc25a3df7472de
During upgrade lib/libOpenCL.so* files/symlinks are not
removed.
This change will remove those symlinks as part of the upgrade
package activity.
Signed-off-by: Saravanan Solaiyappan <saravanan.solaiyappan@amd.com>
Change-Id: Ia12e89e857b358101f93a6044fe090aa7acafbe0
If GlobalMemCacheLine reported is 0, runtime may run into an
infinite loop as the KernelSegmentAlignment is chosen as size of the
cache line.
Change-Id: Ide547940cc0407f16fab10ee210b4fd3ae4eaafc
HIP apps running on CUDA 11.0 are failing due to some graph APIs/enums
added only in CUDA 11.1 or CUDA 11.3
Change-Id: I0d32b412cb76c42c7b3a9c612d750990f9e89908
CPACK_PACKAGING_INSTALL_PREFIX is defined in the wrong location.
Signed-off-by: Jeremy Newton <Jeremy.Newton@amd.com>
Change-Id: Id8ce6beb1d18878c9132cab335cb372a53cc0ee8
RUNPATH was not changed to correct ROCM_RPATH
Packaged proper binaries after RUNPATH correction
Signed-off-by: Ashutosh Mishra <Ashutosh.Mishra@amd.com>
Change-Id: I97443cc0d6cb1ecf7352e3c83ae0fe38a8f339bf
OCL2.2 requires SPIR-V and runtime doesn't support it.
Make sure PAL backend doesn't report any SPIR-V support.
Change-Id: I8d179069674205b54f7d20d149bcb675bee5cdb0
Metadata in Codeobject version 5 is the extension of CO3 and CO4.
Add the detection of the new fields and program them in
the setup of the kernel arguments.
Change-Id: I27e58df77320ad00f4f16d35912668db803826af
German Andryeyev
Saleel Kudchadker
Jeremy Newton
Julia Jiang
Siu Chi Chan
Saravanan Solaiyappan
Chauncey Hui
haoyuan2
Payam
Sarbojit Sarkar
Satyanvesh Dittakavi
Ajay
Sourabh Betigeri
Ashutosh Mishra