| Summary: | mythtv compile fails with gcc-4.4.2 | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Robert Bradbury <robert.bradbury> |
| Component: | Current packages | Assignee: | Gentoo Linux bug wranglers <bug-wranglers> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | normal | ||
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| URL: | N/A | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: | emerge --info | ||
|
Description
Robert Bradbury
2009-11-19 16:35:10 UTC
Created attachment 210633 [details]
emerge --info
Shows I'm using gcc-4.4.2.
Nothing to do with gcc. It's the recent ffmpeg sources hating pic safety on x86. Rafal, I'd agree with your statement possibly regarding the previous h264 errors (I did at one point browse the Gentoo bug report involving the PIC problems) but I'm reluctant to say that the flacdsp_mmx.c error falls into the ffmpeg bin. First of all flac is a sound encoder not a video encoder. Second of all mmx is a standard feature of all x86 processors. So there is the issue of *what* C compiler and/or libraries is mythtv being released with. If one is writing assembler into the C code which is supposed to work on a specific architecture then one is presumably testing it with *some* compiler (be it gcc or some other) so there appears to be some disconnect from the mythtv developers to the gentoo repackagers (unless the mythtv developers have basically gone to a 64 bit architecture and given up on the 32 bit architectures -- which would presumably be unfortunate given the hundreds of millions of x86 machines in the world). Well, take a look at bug 290741: while it's not quite the same and I didn't exactly looked at mythtv sources, still certain function names seem suspiciously similar - perhaps it's not directly ffmpeg related, maybe mythtv based something on ffmeg code, still it fails the same way. *** This bug has been marked as a duplicate of bug 292421 *** |