This bug was fixed in the package glibc - 2.31-0ubuntu9.9
---
glibc (2.31-0ubuntu9.9) focal; urgency=medium
* Disable testsuite on riscv64. It is failing maths tests intermittently in
ways that cannot be a glibc regression and is disabled in later series
anyway.
glibc
Validated glibc from focal-proposed according to test case from
description:
halves@glibc-zen:~$ ./test_memcpy64 32
32 MB = 1.222535 ms
-Compare match (should be zero): 0
halves@glibc-zen:~$ dpkg -l | grep libc-bin
ii libc-bin 2.31-0ubuntu9.9amd64
Hello Heitor, or anyone else affected,
Accepted glibc into focal-proposed. The package will build now and be
available at https://launchpad.net/ubuntu/+source/glibc/2.31-0ubuntu9.5
in a few hours, and then in the -proposed repository.
Please help us by testing this new package. See
Validated glibc from -proposed according to test case from description:
halves@focal-proposed:~$ grep -m1 "model name" /proc/cpuinfo
model name : AMD Ryzen 7 3700X 8-Core Processor
halves@focal-proposed:~$ ./test_memcpy64 32
32 MB = 1.340379 ms
-Compare match (should be zero): 0
Hello Heitor, or anyone else affected,
Accepted glibc into focal-proposed. The package will build now and be
available at https://launchpad.net/ubuntu/+source/glibc/2.31-0ubuntu9.4
in a few hours, and then in the -proposed repository.
Please help us by testing this new package. See
** Description changed:
[Impact]
On AMD Zen systems, memcpy() calls see a heavy performance regression in
Focal and Groovy, due to the way __x86_non_temporal_threshold is calculated.
Before 'glibc-2.33~455', cache values were calculated taking into
consideration the number of hardware
It's been some time since the original benchmarks, so I'm repeating the
test from the description. I haven't used hyperfine for the comparisons
below, so they won't have the same statistical reliability but should
nevertheless be sufficient for validation.
Binaries have been compiled as below:
$
I'm unsubscribing sts-sponsors for now, please feel free to resubscribe
once this is ready for review/spsonsoring to focal.
Note that my understanding on this is that due to the added complexity
of the release and then revert of glibc in focal, due to bug 1914044,
any further upgrades to glibc
There's currently a pending SRU for glibc on Groovy, which I based my
debdiff on. Although that SRU has already cleared the 7 day grace
period, there are still a few pending bugs for verification which we
would need to clear before being able to upload a fix for this one.
Since Groovy is going
** Description changed:
[Impact]
On AMD Zen systems, memcpy() calls see a heavy performance regression in
Focal and Groovy, due to the way __x86_non_temporal_threshold is calculated.
Before 'glibc-2.33~455', cache values were calculated taking into
consideration the number of hardware
** Patch added: "lp1928508-groovy-v2.debdiff"
https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1928508/+attachment/5503158/+files/lp1928508-groovy-v2.debdiff
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
** Patch removed: "lp1928508-focal.debdiff"
https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1928508/+attachment/5502960/+files/lp1928508-focal.debdiff
** Patch removed: "lp1928508-groovy.debdiff"
** Patch added: "lp1928508-groovy.debdiff"
https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1928508/+attachment/5502961/+files/lp1928508-groovy.debdiff
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
** Patch added: "lp1928508-focal.debdiff"
https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1928508/+attachment/5502960/+files/lp1928508-focal.debdiff
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
I've ran the same test on an Intel system, to ensure we aren't
introducing any regressions there. Besides basic smoke tests, the
benchmarks from the description showed that the performance on Intel is
not significantly affected by this patch.
halves@rotom:~$ head -n5 /proc/cpuinfo
processor
Below is the same benchmark from the case description, validating the
performance regression on Groovy. Patched glibc builds for both Focal
and Groovy amd64/i386 are available at ppa:halves/lp1928508-test [0].
$ hyperfine -n groovy-2.32-0ubuntu3 'lxc exec groovy ./test_memcpy64 32' -n
** Description changed:
[Impact]
On AMD Zen systems, memcpy() calls see a heavy performance regression in
Focal and Groovy, due to the way __x86_non_temporal_threshold is calculated.
- Before 'glibc-2.33~455', cache values were calculated taking into
consideration the number of hardware
** Attachment added: "test_memcpy.c"
https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1928508/+attachment/5497631/+files/test_memcpy.c
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1928508
18 matches
Mail list logo