test/kpb: fix tests[1] array out of bounds#2591
test/kpb: fix tests[1] array out of bounds#2591lgirdwood merged 1 commit intothesofproject:masterfrom
Conversation
As reported by gcc. Fixes: 0b279e0. This is a partial, one-line revert of that commit that decreased the size of this array for no obvious reason. Signed-off-by: Marc Herbert <marc.herbert@intel.com>
jajanusz
left a comment
There was a problem hiding this comment.
Weird that it worked at all, cos it was randomly modyfing stack values
|
Jenkins CI known issues. |
|
@jajanusz Btw, do you know if all your team are on GCC 9.x now ? We had some similar fixes last week due to folks updating compiler. |
I doubt anyone is using it here. We mostly have gcc 7/8 on our setups. |
|
@lgirdwood this PR broke UTs in CI, probably you 'fixed' some dirty hack for another issue ;) |
|
@lgirdwood @marc-hb UTs are failing with this patch. The second test should be commented-out/removed to be consistent with the original change instead of being reactivated with no result verification. |
|
@mmaka1 I just realized "Internal Intel CI System/merge/build — Unit tests failed" (a.k.a. "quickbuild") was not reported at all here, any idea why? They were reported in (unrelated) PR #2581 so I took them for granted.
Wait... is Cmocka using
This memory corruption issue was reported by xtensa-apl-elf-gcc (crosstool-NG 1.23.0.440-3b8c) 7.3.0 built by merely following sof-docs + a couple test/cmocka cmake hacks. |
Yes it is, but it's not correct way to 'disable' test. I removed this test in another PR, author of KPB UTs will fix them in the future. |
Unfortunately there is no static complete list of CI items that would contain 'not started yet' items (this was the case I guess). I hope it is possible to create one. I am confused too trying to guess the overall status when looking at the PRs with CI in progress. |
|
As reported by gcc.
Fixes: 0b279e0. This is a partial, one-line revert of that commit
that decreased the size of this array for no obvious reason.
Signed-off-by: Marc Herbert marc.herbert@intel.com