Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ImathTest fails for x86 32-bit platform #346

Closed
waebbl opened this issue Oct 8, 2018 · 5 comments
Closed

ImathTest fails for x86 32-bit platform #346

waebbl opened this issue Oct 8, 2018 · 5 comments
Labels
Bug A bug in the source code Needs Discussion To be discussed in the technical steering committee
Milestone

Comments

@waebbl
Copy link
Contributor

waebbl commented Oct 8, 2018

Hi, I'm co-maintainer for the openexr related packages on Gentoo Linux.

I'm using a multilib platform, allowing 64-bit and 32-bit applications to run in parallel on a x86_64 and using gcc version 7.3.0. IlmBase compiles just fine for both arches, but when I run the testsuite it finishes successfully for 64-bit, but fails for 32-bit. I also tested it in a 32-bit only chroot enviroment with the same result.

It errors on ImathTest with a core dump with

/tmp/portage/media-libs/ilmbase-2.3.0/work/ilmbase-2.3.0/test-driver: line 95: 26717 Aborted                 (core dumped) "$@" > $log_file 2>&1
FAIL: ImathTest

test-suite.log contains the info

ImathTest: /tmp/portage/media-libs/ilmbase-2.3.0/work/ilmbase-2.3.0/ImathTest/testBoxAlgo.cpp:892: void {anonymous}::boxMatrixTransform(): Assertion `b21 == b2' failed.

Any idea, what's the cause for this failure?
TIA

@malaterre
Copy link

The title of the issue is a bit misleading. It should be refer to x86 arch instead (x87 double rounding issue). This issue is not visible for instance on ppc, mips or arm (although 32bits architectures).

@malaterre
Copy link

Here is a silly patch:

Comments welcome

@waebbl waebbl changed the title ImathTest fails for 32-bit platform ImathTest fails for x86 32-bit platform Nov 30, 2018
@cary-ilm cary-ilm added the Bug A bug in the source code label Jun 13, 2019
@cary-ilm cary-ilm added this to the Backlog milestone Feb 7, 2020
@cary-ilm cary-ilm added the Needs Discussion To be discussed in the technical steering committee label Feb 7, 2020
@waebbl
Copy link
Contributor Author

waebbl commented Mar 3, 2020

Sorry, I totally lost track of this issue.
We're in the process on moving to 2.4 release on Gentoo, where I didn't encounter this problem anymore. On my tests so far, all test cases succeeded.
Therefore, from my point of view, this issue can be closed.

@cary-ilm
Copy link
Member

cary-ilm commented Mar 5, 2020

Closing the issue for now, feel free to re-open or file a new issue if you need further help.

@cary-ilm cary-ilm closed this as completed Mar 5, 2020
@waebbl
Copy link
Contributor Author

waebbl commented Jul 3, 2020

Here is a silly patch:

Comments welcome

This issue popped up again on x86 platform. However, by adding -ffloat-store from @malaterre's patch to CPPFLAGS for this platform solves it.
Thanks for the patch.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Bug A bug in the source code Needs Discussion To be discussed in the technical steering committee
Projects
None yet
Development

No branches or pull requests

3 participants