| 1 |
Samir Mishra posted <444A8470.3060501@×××.ae>, excerpted below, on Sat, |
| 2 |
22 Apr 2006 23:30:56 +0400: |
| 3 |
|
| 4 |
> Thought I'd post this, in case someone else comes across this problem, |
| 5 |
> or there's a solution. |
| 6 |
> |
| 7 |
> I'm unable to compile libmpeg3 with gcc-4.1.0, though I have no problems |
| 8 |
> with gcc-3.4.5 (hardened). |
| 9 |
|
| 10 |
No problems here. I had libmpeg3-1.5.2-r3 (latest ~amd64) already merged, |
| 11 |
probably with gcc-4.1.0 tho I wasn't sure so I remerged it (gcc-4.1.0 is |
| 12 |
my default), no problem at all. |
| 13 |
|
| 14 |
There's no USE flags for it, so that couldn't be it. |
| 15 |
|
| 16 |
Are you running ~amd64 or stable amd64 for your general system? I'd not |
| 17 |
suggest trying to run stable with gcc-4.1, if you are trying that. |
| 18 |
Keyworded stable ebuilds likely haven't had the gcc-4 and 4.1 fixes |
| 19 |
applied that many ~arch ebuilds will have had, as gcc-4.1 isn't even |
| 20 |
unmasked yet, and any fixes will have likely only been applied to the |
| 21 |
newest ~arch ebuilds. |
| 22 |
|
| 23 |
It could also be CFLAGS/CXXFLAGS. Did you try trimming those to the usual |
| 24 |
recommended "-march=k8 -pipe -O2"? (Of course, replace that k8 with |
| 25 |
whatever em64t uses if you have an Intel. FWIW, k8, athlon64, opteron, |
| 26 |
etc, are all aliased to the same thing internal to gcc, so that doesn't |
| 27 |
matter, but the em64t one would.) |
| 28 |
|
| 29 |
For reference, here's mine, which as I said work: |
| 30 |
CFLAGS="-march=k8 -Os -pipe -fomit-frame-pointer -frename-registers |
| 31 |
-funit-at-a-time -fweb -freorder-blocks -freorder-blocks-and-partition |
| 32 |
-ftree-pre -fmerge-all-constants" |
| 33 |
|
| 34 |
CXXFLAGS is the same, only minus the -freorder-blocks-and-partition flag, |
| 35 |
since it seems to be disabled for most C++ anyway, and I got tired of the |
| 36 |
warnings and occasional failed configure test due to them. |
| 37 |
|
| 38 |
> Also, to get libdts to install, I had to first install autotoolset, |
| 39 |
> though it's not listed as a dependency. |
| 40 |
|
| 41 |
Likewise, libdts-0.0.2-r3 merged, emerge pretend says autotoolset not |
| 42 |
merged. A quick remerge of libdts was uneventful. USE="oss -debug". |
| 43 |
|
| 44 |
Note however that I'm running without FEATURES=test. If you have test |
| 45 |
toggled on and the problem is occurring during the test phase, that's |
| 46 |
likely the difference there. I'd suggest filing a bug if one's not filed |
| 47 |
yet. |
| 48 |
|
| 49 |
I don't believe they want gcc-4.1 bugs yet unless you have patches, so I'd |
| 50 |
not recommend filing bugs you suspect might be 4.1 related, unless you |
| 51 |
know exactly what's wrong and can suggest a fix. |
| 52 |
|
| 53 |
So... you know both of those problems don't happen to everyone. If you |
| 54 |
post the errors, maybe I can compare to my logs and see what's going on. |
| 55 |
|
| 56 |
-- |
| 57 |
Duncan - List replies preferred. No HTML msgs. |
| 58 |
"Every nonfree program has a lord, a master -- |
| 59 |
and if you use the program, he is your master." Richard Stallman in |
| 60 |
http://www.linuxdevcenter.com/pub/a/linux/2004/12/22/rms_interview.html |
| 61 |
|
| 62 |
|
| 63 |
-- |
| 64 |
gentoo-amd64@g.o mailing list |
Archives