Summary: | dev-lang/v8-3.0.12.30 - tests fail & dysfunction on armv7a-hardfloat-linux-gnueabi | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Fabian Köster <gentoo> |
Component: | [OLD] Unspecified | Assignee: | Chromium Project <chromium> |
Status: | RESOLVED UPSTREAM | ||
Severity: | normal | CC: | arm, siarhei.siamashka |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | ARM | ||
OS: | Linux | ||
URL: | http://code.google.com/p/v8/issues/detail?id=1301 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | build.log with failing tests |
Description
Fabian Köster
2011-03-22 11:02:59 UTC
Created attachment 266827 [details]
build.log with failing tests
CC-ing ARM arch team, I have no idea how to fix those issues. JIT definitely needs to be updated to support hardfloat calling conventions. And this is probably better to be done in upstream because the patches may be not so trivial. For example, Mozilla added hardfloat support to NanoJIT for Firefox4 via the following bug: https://bugzilla.mozilla.org/show_bug.cgi?id=602834 ARM hardfloat is still not ready for casual users and it has some problems. JIT is one of the sources of such problems. And also it is much harder to find usable closed source binary drivers for hardfloat ABI, if somebody wants hardware accelerated OpenGL ES for example. I now cross-checked on an equal system with softfloat-toolchain and chromium (10.0.648.204) works perfectly there. I think I will stay on softfloat for now because it causes less trouble (e.g. binary drivers work). I reported a bug upstream [1] so you may close this bug as RESOLVED UPSTREAM now. [1] http://code.google.com/p/v8/issues/detail?id=1301 Thank you for testing and forwarding the bug upstream. |