Add pointer overloads for Avx2.BroadcastScalarToVector128#20055
Conversation
|
After this PR and #19420 gets merged, I will submit a PR to implement all the remaining AVX2 intrinsic in JIT. |
fiigii
force-pushed
the
addbroadcast
branch
from
5d580f5 to
562d663
Compare
|
Added comments and disabled tests of |
| /// __m128i _mm_broadcastw_epi16 (__m128i a) | ||
| /// VPBROADCASTW xmm, m16 | ||
| /// The above native signature does not directly correspond to the managed signature. | ||
|
|
There was a problem hiding this comment.
nit: stray newlines in the middle of the doc comments for most of these.
fiigii
force-pushed
the
addbroadcast
branch
from
562d663 to
664c3d4
Compare
| /// __m128i _mm_broadcastb_epi8 (__m128i a) | ||
| /// VPBROADCASTB xmm, m8 | ||
| /// The above native signature does not directly correspond to the managed signature. | ||
|
|
|
Thanks for the work @fiigii. Feel free to tag me when the tests complete and I'll get this merged. |
fiigii
force-pushed
the
addbroadcast
branch
from
664c3d4 to
df1712e
Compare
fiigii
force-pushed
the
addbroadcast
branch
from
19bb026 to
cb5d7f4
Compare
|
@dotnet-bot test this please |
|
CI is still stuck... Can we merge this and #19420? I have submitted corefx PR. |
|
Many PRs are stuck on testing, but I'm not comfortable with the idea that we should just merge without completing at least the xarch testing that's in the main CI set. |
|
Merging, Two of the pending jobs actually completed. The other was an out of disk space issue, but a similar job (testing additional functionality) succeeded. For the additional failures, they are also unrelated CI failures and had been previously passing (they were rerun due to a new commit which cleaned up some code comments). |
|
@CarolEidt, saw your comment after I hit merge. The xarch jobs were all passing, CI just didn't update. All the jobs had also been passing before the code comment cleanup I asked @fiigii to do, so the minor cleanup that did occur shouldn't have caused any additional regressions in the remaining jobs (which were mostly the perf ones). |
|
@tannergooding - thanks for the clarification, and for the work of tracking down job status. |
4 participants
close https://github.com/dotnet/corefx/issues/30266