10.0.1-rc2 release has been tagged

Hi,

I've tagged the 10.0.1-rc2 release, please test the release and report any issues.

Thanks,
Tom

Uploaded Xubuntu 20.04. Added ‘xubuntu-20.04-’ to the beginning of the tar.xz file to avoid possible overwrite clash with another upload.

sha256sum xubuntu-20.04-clang+llvm-10.0.1-rc2-x86_64-unknown-linux-gnu.tar.xz
db610bb038c010badb97a9b60492af2a8d361edf2072d2f81963d2af4b578bcc

scp -i ~/.ssh/id_rsa_llvm.pub xubuntu-20.04-clang+llvm-10.0.1-rc2-x86_64-unknown-linux-gnu.tar.xz xubuntu-20.04-clang+llvm-10.0.1-rc2-x86_64-unknown-linux-gnu.tar.xz 99% 330MB 267.9KB/s 00:00 ETAscp: /home/testers/xubuntu-20.04-clang+llvm-10.0.1-rc2-x86_64-unknown-linux-gnu.tar.xz: xubuntu-20.04-clang+llvm-10.0.1-rc2-x86_64-unknown-linux-gnu.tar.xz 100% 330MB 208.4KB/s 26:59

IL: libomp :: lock/omp_init_lock.c (67443 of 69960)
******************** TEST ‘libomp :: lock/omp_init_lock.c’ FAILED ********************
Script:

Windows is ready:

$ sha1sum LLVM-10.0.1-rc2*.exe
79dd8f3e9e509709aaf3ab0226f1f84ad8d265e1 LLVM-10.0.1-rc2-win32.exe
4b7a4ccab4b6fef1ba85c7035a51207a89c4dd67 LLVM-10.0.1-rc2-win64.exe

It was built with the attached batch file (which is based on
llvm/utils/release/build_llvm_package.bat).

There was one test failure: llvm/test/CodeGen/X86/asm-modifier2.ll
That was fixed on trunk, but the fix didn't reach the branch. I've
commented on PR46223.

build_llvm_1001-rc2.bat|attachment (4.97 KB)

So far one regression in LLVM tests, apparently caused by incomplete asm
fix backport. FWICS additional backport was requested already:

https://bugs.llvm.org/show_bug.cgi?id=46223

For this rc, I used four patches, which are attached.

Main results on amd64-freebsd11:

  Expected Passes : 68020
  Passes With Retry : 1
  Expected Failures : 265
  Unsupported Tests : 4654
  Unresolved Tests : 1
  Unexpected Passes : 6
  Unexpected Failures: 542
  Individual Timeouts: 13

Test suite results on amd64-freebsd11:

  Expected Passes : 2398
  Unexpected Failures: 3

Main results on i386-freebsd11:

  Expected Passes : 65077
  Passes With Retry : 2
  Expected Failures : 248
  Unsupported Tests : 3083
  Unexpected Passes : 6
  Unexpected Failures: 225
  Individual Timeouts: 10

Uploaded:
SHA256 (clang+llvm-10.0.1-rc2-amd64-unknown-freebsd11.tar.xz) = e81a26433c45b3a477686a8a361c9fb2dc5772f91b0841ebedeb01e9bebecac9
SHA256 (clang+llvm-10.0.1-rc2-i386-unknown-freebsd11.tar.xz) = 3ad9341b2548f9a48536e4068823c36af9ff32dc13c09f07493bd862ac4c79ba

-Dimitry

fix-clang-1.diff (447 Bytes)

fix-compiler-rt-1.diff (890 Bytes)

fix-libcxx-1.diff (987 Bytes)

fix-test-suite-1.diff (552 Bytes)

Hello,

I finished testing llvm-10.0.1 RC2 on Power PC 64bit Little Endian Ubuntu 16.04 machine and have uploaded the binary from IBM.

The only regression was the lit failure of llvm/test/CodeGen/X86/asm-modifier2.ll and this has also been mentioned by Hans’ email.

The sha1 file is attached.

Thanks,
Anil Mahmud.

clang+llvm-10.0.1-rc2-powerpc64le-linux-ubuntu-16.04.sha1 (102 Bytes)

Hello,

I finished testing llvm-10.0.1 RC2 on Power PC 64bit Little Endian Red Hat 7.4 machine and have uploaded the binary from IBM.
The only regression was the lit failure of llvm/test/CodeGen/X86/asm-modifier2.ll and this has been mentioned in previous emails.

The sha1 file is attached.

Thanks,
Anil Mahmud

clang+llvm-10.0.1-rc2-powerpc64le-linux-rhel-7.4.sha1 (98 Bytes)

Tom,

Uploaded clang+llvm-10.0.1-rc2-x86_64-apple-darwin.tar.xz with SHA256:
014cd349ffc5562205393072b2e99dcab9ad178a4c0d8267d5e42301aed67bcf

Nothing stands out in the tests the failing ones are:

FAIL: LLVM :: Bindings/Go/go.test (25257 of 67107)
FAIL: ThreadSanitizer-x86_64 :: Darwin/norace-objcxx-run-time.mm
(57277 of 67107)
FAIL: libunwind :: signal_frame.pass.cpp (64811 of 67107)

Thanks,
Tobias

Hi,

Uploaded ARM & AArch64:
de9ea1211c7ae33de35140343833c5fe7a9574de2d089e51f1d86588b21c7e53
clang+llvm-10.0.1-rc2-aarch64-linux-gnu.tar.xz
7fd69f1922823fcd1904b70b5d0dd0b5f0084d7dd84a37886f874c486963b01b
clang+llvm-10.0.1-rc2-armv7a-linux-gnueabihf.tar.xz

Both fail llvm/test/CodeGen/X86/asm-modifier2.ll as mentioned by other
testers. Otherwise no new failures, and I can confirm that PR46092 is
fixed for armhf.

Thanks,
Diana

Looks good to me in Debian experimental!

Cheers,
Sylvestre

Hi,
finally got around to updating openmandriva -- a few bugs from 10.0 remain (nothing too serious, mostly mismatches between libgcc and compiler-rt), but no regressions so far.

ttyl
bero