--- Josh Coalson <xflac@yahoo.com> wrote:> > -- Miroslav Lichvar <lichvarm@phoenix.inf.upol.cz> wrote: > > On Tue, Dec 17, 2002 at 01:01:08PM -0800, Josh Coalson wrote: > > > --- Miroslav Lichvar <lichvarm@phoenix.inf.upol.cz> wrote: > > > > Ok, what about enabling the 3dnow function in libFLAC by > default? > > > > I think time has shown the function is bugfree... :) > > > > > > Yeah, I just haven't done it because I don't remember hearing > > > feedback from others about using it (or maybe I just forgot... > > > has anyone else built flac 1.0.4 with ./configure --enable-3dnow > ?) > > > Also I don't have an AMD to test it on. > > > > > > If the current test suite (CVS as of last night) works for you > > > though it is probably OK. > > > > I have a problem with make check: > > ... > > make[3]: Entering directory > `/home/lichvarm/tmp/flac/flac-test/test' > > Makefile:310: *** Recursive variable `TESTS_ENVIRONMENT' references > > itself (eventually). Stop. > > make[3]: Leaving directory `/home/lichvarm/tmp/flac/flac-test/test' > > make[2]: *** [check-am] Error 2 > > ... > > Oops, I'll fix that soon.fixed in CVS.> > So i ran "make -f Makefile.lite release" in test directory. All > have > > passed (except test_bins.sh, there was no test_files/bins/ > > directory). > > OK, I'll make it the default in 1.0.5now the default. Josh __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com
On Mon, Dec 23, 2002 at 10:20:38PM -0800, Josh Coalson wrote:> --- Josh Coalson <xflac@yahoo.com> wrote: > > > > OK, I'll make it the default in 1.0.5 > > now the default.I did an extra check and I found stack handling in the function isn't 100% clean, I can imagine possible segfault. Please apply attached patch. Thanks. -- Miroslav Lichvar -------------- next part -------------- --- src/libFLAC/ia32/lpc_asm.nasm.orig 2002-01-26 19:05:12.000000000 +0100 +++ src/libFLAC/ia32/lpc_asm.nasm 2002-12-26 02:16:12.000000000 +0100 @@ -619,6 +619,8 @@ mov esi, [ebp + 20] mov edi, [ebp + 24] mov edx, [ebp + 28] + inc edx + and edx, byte -2 mov eax, edx neg eax and esp, byte -8 @@ -639,7 +641,7 @@ cmp esi, ecx mov eax, esi ja short .loop2_pre - align 16 ;8 nops + align 16 ;4 nops .loop1_i: movd mm0, [eax] movd mm2, [eax + 4] @@ -706,6 +708,7 @@ jnz short .loop2_i mov edi, [ebp + 32] + mov edx, [ebp + 28] .loop3: dec edx mov eax, [esp + 4 * edx]
On Thu, Dec 26, 2002 at 03:11:03AM +0100, Miroslav Lichvar wrote:> On Mon, Dec 23, 2002 at 10:20:38PM -0800, Josh Coalson wrote: > > --- Josh Coalson <xflac@yahoo.com> wrote: > > > > > > OK, I'll make it the default in 1.0.5 > > > > now the default. > > I did an extra check and I found stack handling in the function isn't > 100% clean, I can imagine possible segfault. Please apply attached > patch. Thanks.Ehm, I take it back. The function is absolutely ok and the patch is useless. Sorry for the confusion. :) -- Miroslav Lichvar