[GHA] Uplift Linux GPU RT version to 24.45.31740.9 #16312
Conversation
bb-sycl
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
bb-sycl
temporarily deployed
to
WindowsCILock
GitHub Actions
Inactive
|
@intel/dpcpp-devops-reviewers, ping. |
|
Sorry, forgot to assign myself. the IGC CI was totally busted, I had to fix it myself and now it's somewhat working, will work on this. |
Signed-off-by: Sarnie, Nick <[email protected]>
Signed-off-by: Sarnie, Nick <[email protected]>
…r-linux-24.45.31740.9
Signed-off-by: Sarnie, Nick <[email protected]>
Signed-off-by: Sarnie, Nick <[email protected]>
Signed-off-by: Sarnie, Nick <[email protected]>
Signed-off-by: Sarnie, Nick <[email protected]>
|
@intel/dpcpp-sanitizers-review Ping, needed to disable some tests that started failing in the GPU driver, please take a look, thanks! |
|
@uditagarwal97 Hey do you mind reviewing the devops changes here? The new stable GPU driver requires Ubuntu 24.04, so we switch to that everywhere we use the driver and update the install driver script to expect 24.04 as well as deal with the current upstream dpkg issues that previously existed in igc-dev but now also exist in stable, it's known upstream already. Thx! |
There was a problem hiding this comment.
@intel/dpcpp-sanitizers-review Ping, needed to disable some tests that started failing in the GPU driver, please take a look, thanks!
@sarnex I'm a bit confused, this is the only change I see since my last review, how does xfailing in CPU relate to problems in the GPU driver?
There was a problem hiding this comment.
yeah youre right this makes no sense, let me remove it try to remember why i added it, thanks
There was a problem hiding this comment.
ok it took me forever but i investigated it, basically the test calls the gpu driver manually even when the target device is cpu , that's why it's failing, so im gonna add the XFAIL back, idk if thats intentional or not but ill let the test owner investigate
There was a problem hiding this comment.
Is that expected? Should we rewrite the test not to do that?
There was a problem hiding this comment.
no idea, theres a GH issue tracking the XFAIL so i'll let the issue owner make a decision
Signed-off-by: Sarnie, Nick <[email protected]>
…r-linux-24.45.31740.9
Signed-off-by: Sarnie, Nick <[email protected]>
Signed-off-by: Sarnie, Nick <[email protected]>
Signed-off-by: Sarnie, Nick <[email protected]>
6 participants
Scheduled drivers uplift