Summary: | =media-libs/xine-lib-1.2.6 - eq.c: In function 'process_MMX': eq.c:52:3: error: 'asm' operand has impossible constraints | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Bertrand Jacquin <bertrand> |
Component: | Hardened | Assignee: | Gentoo Media-video project <media-video> |
Status: | CONFIRMED --- | ||
Severity: | normal | CC: | bertrand, hardened, quantheory |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
build.log
emerge --info build.log emerge --info build.log (IUSE+=mmx) |
Description
Bertrand Jacquin
2014-02-01 17:03:04 UTC
Created attachment 369276 [details]
build.log
Created attachment 369278 [details]
emerge --info
This only append with gcc version 4.8.2 (Gentoo Hardened 4.8.2 p1.3r1, pie-0.5.8r1). Workaround is to use gcc version 4.7.3 (Gentoo Hardened 4.7.3-r1 p1.4, pie-0.5.5) xine-lib (1.2.5) 2014-04-08 ChangeLog says Build fixes. Does xine-lib-1.2.5 compile with gcc-4.8? =media-libs/xine-lib-1.2.4 removed from Portage as unused, in favour of 1.2.5 Does 1.2.5 still fail? (In reply to Samuli Suominen from comment #5) > =media-libs/xine-lib-1.2.4 removed from Portage as unused, in favour of 1.2.5 > > Does 1.2.5 still fail? Still :/ Do you need an updated build.log ? (In reply to Bertrand Jacquin from comment #6) > (In reply to Samuli Suominen from comment #5) > > =media-libs/xine-lib-1.2.4 removed from Portage as unused, in favour of 1.2.5 > > > > Does 1.2.5 still fail? > > Still :/ Do you need an updated build.log ? I suppose it wouldn't hurt Just trying to figure out the xine-lib/xine-ui/gxine bug situation, what are still relavent and what are not, they've been neglected for a while, sorry about that Created attachment 379428 [details]
build.log
Created attachment 379430 [details]
emerge --info
ssuominen@null /tmp $ diff -u xine-lib-1.2.5/src/post/planar/eq.c xine-lib-1.2.6/src/post/planar/eq.c ssuominen@null /tmp $ Hrr. Just pointing out 1.2.6 is in tree and I expect this still to be unfixed. I haven't been able to reproduce this yet... I have a feeling this is related to hardened. I believe that this is related to -fstack-check being enabled on hardened. See Bug 499996, and Bug 506206. (In reply to Sean Santos from comment #12) > I believe that this is related to -fstack-check being enabled on hardened. > See Bug 499996, and Bug 506206. Yes, because pic/pie takes away one of the registers. To be certain, can the original poster switch to the vanilla gcc toolchain (using gcc-config) and try again. This should not be blocking gcc-4.8 (In reply to Anthony Basile from comment #13) > (In reply to Sean Santos from comment #12) > > I believe that this is related to -fstack-check being enabled on hardened. > > See Bug 499996, and Bug 506206. > > Yes, because pic/pie takes away one of the registers. To be certain, can > the original poster switch to the vanilla gcc toolchain (using gcc-config) > and try again. This should not be blocking gcc-4.8 ACK, build is OK with i686-pc-linux-gnu-4.8.3-vanilla You could try this: First add "mmx" to IUSE: IUSE="mmx" And this line to end of src_prepare() function in the ebuild: use mmx || { sed -i -e 's:HAVE_MMX:HARDENED_AND_MMX_FAILS:g' configure || die; } Then try to build with USE="-mmx" As in, we could add this USE flag to control building MMX features and then mask it in hardened profiles... Created attachment 382784 [details] build.log (IUSE+=mmx) (In reply to Samuli Suominen from comment #15) > You could try this: > > First add "mmx" to IUSE: > > IUSE="mmx" > > And this line to end of src_prepare() function in the ebuild: > > use mmx || { sed -i -e 's:HAVE_MMX:HARDENED_AND_MMX_FAILS:g' configure || > die; } > > Then try to build with USE="-mmx" > > As in, we could add this USE flag to control building MMX features and then > mask it in hardened profiles... I've just try it, and I have another failure libtool: compile: i686-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I../../../include -I../../.. -I../../../include -I../../../include -I../../../src -I../../../src/xine-engine -I../../../src/xine-engine -I../../../src/xine-utils -I../../../src/input -I../../../src/input -I../../../lib -I../../../lib -D_FILE_OFFSET_BITS=64 -DNDEBUG -D_REENTRANT -DXINE_COMPILE -fvisibility=hidden -pipe -Wall -Wformat=2 -Wno-format-zero-length -Wmissing-format-attribute -Werror-implicit-function-declaration -Wstrict-aliasing=2 -Wchar-subscripts -Wmissing-declarations -Wmissing-prototypes -Wwrite-strings -Wpointer-arith -g -march=native -O2 -pipe -fomit-frame-pointer -fomit-frame-pointer -Wall -Wchar-subscripts -Wnested-externs -Wcast-align -Wmissing-declarations -Wmissing-prototypes -Wmissing-format-attribute -Wno-pointer-sign -Wformat=2 -Wno-format-zero-length -Wformat-security -Wstrict-aliasing=2 -Werror=implicit-function-declaration -c invert.c -fPIC -DPIC -o .libs/xineplug_post_planar_la-invert.o eq.c: In function 'process_MMX': eq.c:52:3: error: 'asm' operand has impossible constraints asm volatile ( ^ Makefile:727: recipe for target 'libpost_planar_asm_la-eq.lo' failed make[3]: *** [libpost_planar_asm_la-eq.lo] Error 1 make[3]: *** Waiting for unfinished jobs.... noise.c: In function 'lineNoiseAvg_MMX': noise.c:232:5: error: 'asm' operand has impossible constraints asm volatile( ^ Makefile:741: recipe for target 'libpost_planar_asm_la-noise.lo' failed make[3]: *** [libpost_planar_asm_la-noise.lo] Error 1 make[3]: Leaving directory '/var/tmp/portage/media-libs/xine-lib-1.2.6/work/xine-lib-1.2.6/src/post/planar' Makefile:586: recipe for target 'all-recursive' failed make[2]: *** [all-recursive] Error 1 make[2]: Leaving directory '/var/tmp/portage/media-libs/xine-lib-1.2.6/work/xine-lib-1.2.6/src/post' Makefile:604: recipe for target 'all-recursive' failed make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory '/var/tmp/portage/media-libs/xine-lib-1.2.6/work/xine-lib-1.2.6/src' Makefile:680: recipe for target 'all-recursive' failed make: *** [all-recursive] Error 1 * ERROR: media-libs/xine-lib-1.2.6::gentoo failed (compile phase): * emake failed Full log attached |