Tweaked glibc 2.4 + 2.5 + snapshot ebuilds

Message

vipernicus · Post by **vipernicus** » Mon Oct 02, 2006 3:23 pm

nowinter wrote:
nxsty wrote: -Bdirect is obsolete
I thought it is your ebuild, namely 2.17, that makes use of -Bdirect, and the one I actually use (although I also have 2.17.50.0.3). Does what I quoted mean I should make use of -hash-style only perspectively? I must have missed something in this thread.

At the moment it doesn't look like Michael Meeks is going to continue development of bdirect. Prelink and -hashstyles is just as good of an alternative and more featured in my opinion.

nxsty · Post by **nxsty** » Mon Oct 02, 2006 3:31 pm

vipernicus wrote:
nowinter wrote:
nxsty wrote: -Bdirect is obsolete
I thought it is your ebuild, namely 2.17, that makes use of -Bdirect, and the one I actually use (although I also have 2.17.50.0.3). Does what I quoted mean I should make use of -hash-style only perspectively? I must have missed something in this thread.
At the moment it doesn't look like Michael Meeks is going to continue development of bdirect. Prelink and -hashstyles is just as good of an alternative and more featured in my opinion.

I was about to post the same thing but then I checked and found that there actually is a new version of the -Bdirect patch for glibc 2.5.

So with some luck it'll work with --hash-style this time. I'll include it in my next update!

vipernicus · Post by **vipernicus** » Mon Oct 02, 2006 8:31 pm

nxsty wrote:
vipernicus wrote:
nowinter wrote:
nxsty wrote: -Bdirect is obsolete
I thought it is your ebuild, namely 2.17, that makes use of -Bdirect, and the one I actually use (although I also have 2.17.50.0.3). Does what I quoted mean I should make use of -hash-style only perspectively? I must have missed something in this thread.
At the moment it doesn't look like Michael Meeks is going to continue development of bdirect. Prelink and -hashstyles is just as good of an alternative and more featured in my opinion.
I was about to post the same thing but then I checked and found that there actually is a new version of the -Bdirect patch for glibc 2.5. So with some luck it'll work with --hash-style this time. I'll include it in my next update!

What about bdirect with binutils? Last time I talked to Michael Meeks he said that someone else was porting bdirect and hashvals to newer binutils and glibc, but it wasn't to work with hash-style.

7seven7 · Post by **7seven7** » Tue Oct 03, 2006 10:24 am

I have my system with glibc-2.4-r4 and binutils-2.16.93 from nxsty, compiled with this LDFLAGS "-Wl,-O1 -Wl,--sort-common -Wl,-Bdirect -Wl,-hashvals -Wl,-zdynsort -Wl,--as-needed"

I want to drop LDFLAGS "-Wl,-Bdirect -Wl,-hashvals -Wl,-zdynsort" to start use "-Wl, -hash-style=both", with prelink.

Can i just change the LDFLAGS and upgrade to glibc and binutls that are in portage, or is more complex to do the transition?

nxsty · Post by **nxsty** » Tue Oct 03, 2006 11:46 am

7seven7 wrote:I have my system with glibc-2.4-r4 and binutils-2.16.93 from nxsty, compiled with this LDFLAGS "-Wl,-O1 -Wl,--sort-common -Wl,-Bdirect -Wl,-hashvals -Wl,-zdynsort -Wl,--as-needed"

I want to drop LDFLAGS "-Wl,-Bdirect -Wl,-hashvals -Wl,-zdynsort" to start use "-Wl, -hash-style=both", with prelink.

Can i just change the LDFLAGS and upgrade to glibc and binutls that are in portage, or is more complex to do the transition?

There is a guide on the gentoo wiki:
http://gentoo-wiki.com/HOWTO_Hashstyle# ... _hashstyle

7seven7 · Post by **7seven7** » Tue Oct 03, 2006 11:56 am

nxsty wrote:There is a guide on the gentoo wiki:
http://gentoo-wiki.com/HOWTO_Hashstyle# ... _hashstyle

What is the best/recent glibc, the one from portage or from your overlay?

roderick · Post by **roderick** » Tue Oct 03, 2006 6:58 pm

7seven7 wrote:
nxsty wrote:There is a guide on the gentoo wiki:
http://gentoo-wiki.com/HOWTO_Hashstyle# ... _hashstyle
What is the best/recent glibc, the one from portage or from your overlay?

Use the latest binutils in portage.
Use the 2.4-r4 glibc in overlay.

Follow the migration guide as listed above.

If you do this, all should work as expected.

You can try 2.5, though I haven't yet, so I cannot comment on any issues or difficulties you may have. At least with the 2.4 path, you can alway update it after to 2.5 if you want. Just remember to quickpkg your old binutils, glibc prior to making this shift. It may help you recover from a disaster.

nxsty · Post by **nxsty** » Tue Oct 03, 2006 7:27 pm

7seven7 wrote:
nxsty wrote:There is a guide on the gentoo wiki:
http://gentoo-wiki.com/HOWTO_Hashstyle# ... _hashstyle
What is the best/recent glibc, the one from portage or from your overlay?

My glibc 2.5 ebuild isn't much different from the one in portage yet. So either will do. If you want the stable, tested 2.4-r4 you need my overlay.

7seven7 · Post by **7seven7** » Tue Oct 03, 2006 8:23 pm

With your glibc-2.5 the use flag hashstyle is needed?

EDIT: And prelink, is done by portage or has to be done manual?

nxsty · Post by **nxsty** » Thu Oct 05, 2006 3:30 pm

7seven7 wrote:With your glibc-2.5 the use flag hashstyle is needed?

EDIT: And prelink, is done by portage or has to be done manual?

No, the --hash-style patches are included in the upstream release so they're always enabled.

It should be done automatically but I usually run it anyway just to be sure.

nxsty · Post by **nxsty** » Thu Oct 05, 2006 3:32 pm

I updated my overlay with a new snapshot and the latest -Bdirect patch which now works with glibc 2.5 again! I also removed the hashstyle USE-flag now when both the -Bdirect and --hash-style patches can be applied at the same time (though I don't think they can be used at the same time, please test someone.

)

vipernicus · Post by **vipernicus** » Thu Oct 05, 2006 6:38 pm

In order to get the full benefit of your glibc, would I need to rebuild world with it?

nxsty · Post by **nxsty** » Thu Oct 05, 2006 7:07 pm

If you´ve already built your system with --hash-style then probably not.

cruzki · Post by **cruzki** » Fri Oct 06, 2006 9:51 am

i'm try to test if -Bdirect and hash style can co-exists but if I use you binutils-2.17 overlays I have -Bdirect but no -hashstyle an dif I use portage bintutils-2.17.50.5 I have -hashstyle but no -bdirect (I have a C compiler cannot create ejecutable or things like that in both case)

I thing this is because binutils don't have the two patchets at the same time, am I wrong?

PS: sorry for my bad english

vipernicus · Post by **vipernicus** » Fri Oct 06, 2006 2:56 pm

cruzki wrote:i'm try to test if -Bdirect and hash style can co-exists but if I use you binutils-2.17 overlays I have -Bdirect but no -hashstyle an dif I use portage bintutils-2.17.50.5 I have -hashstyle but no -bdirect (I have a C compiler cannot create ejecutable or things like that in both case)

I thing this is because binutils don't have the two patchets at the same time, am I wrong?

PS: sorry for my bad english

You need to have a binutils with the bdirect patches. binutils-2.17.50.05 doesn't have them.

JoKo · Post by **JoKo** » Fri Oct 06, 2006 8:32 pm

nxsty wrote:I updated my overlay with a new snapshot and the latest -Bdirect patch which now works with glibc 2.5 again! I also removed the hashstyle USE-flag now when both the -Bdirect and --hash-style patches can be applied at the same time (though I don't think they can be used at the same time, please test someone. )

I'm using sys-libs/glibc-2.4.90.20060915.

Now there is sys-libs/glibc-2.5.20061005 from your overlay... In order to update and keep the hash-style patches, I just emerge the new version? Also, should I delete the hashstyle USE-flag for good? Any other change to make on the config files?

nxsty · Post by **nxsty** » Fri Oct 06, 2006 9:19 pm

JoKo wrote:
nxsty wrote:I updated my overlay with a new snapshot and the latest -Bdirect patch which now works with glibc 2.5 again! I also removed the hashstyle USE-flag now when both the -Bdirect and --hash-style patches can be applied at the same time (though I don't think they can be used at the same time, please test someone. )
I'm using sys-libs/glibc-2.4.90.20060915.

Now there is sys-libs/glibc-2.5.20061005 from your overlay... In order to update and keep the hash-style patches, I just emerge the new version? Also, should I delete the hashstyle USE-flag for good? Any other change to make on the config files?

You can use 2.5 or the 2.5.20061005 snapshot as hash-style is included in the upstream release now.

cruzki · Post by **cruzki** » Sat Oct 07, 2006 9:42 am

vipernicus wrote:
cruzki wrote:i'm try to test if -Bdirect and hash style can co-exists but if I use you binutils-2.17 overlays I have -Bdirect but no -hashstyle an dif I use portage bintutils-2.17.50.5 I have -hashstyle but no -bdirect (I have a C compiler cannot create ejecutable or things like that in both case)

I thing this is because binutils don't have the two patchets at the same time, am I wrong?

PS: sorry for my bad english
You need to have a binutils with the bdirect patches. binutils-2.17.50.05 doesn't have them.

The one in the nxsty overlay hasn't it? (binutils-2.17 no 2.17.0.5) O.o

nxsty · Post by **nxsty** » Sat Oct 07, 2006 11:58 am

cruzki wrote:
vipernicus wrote:
cruzki wrote:i'm try to test if -Bdirect and hash style can co-exists but if I use you binutils-2.17 overlays I have -Bdirect but no -hashstyle an dif I use portage bintutils-2.17.50.5 I have -hashstyle but no -bdirect (I have a C compiler cannot create ejecutable or things like that in both case)

I thing this is because binutils don't have the two patchets at the same time, am I wrong?

PS: sorry for my bad english
You need to have a binutils with the bdirect patches. binutils-2.17.50.05 doesn't have them.
The one in the nxsty overlay hasn't it? (binutils-2.17 no 2.17.0.5) O.o

I added an ebuild for binutils 2.17.50.0.5 with a bdirect USE-flag and a regression fix, but I couldn´t get it to build when -Bdirect is enabled. Everbody using binutils 2.17.50.0.5 from portage should use this ebuild instead as the regression is pretty serious (binutils bug #3314) Get it from toolchain_overlay.

JoKo · Post by **JoKo** » Sat Oct 07, 2006 4:04 pm

nxsty wrote:
JoKo wrote:
nxsty wrote:I updated my overlay with a new snapshot and the latest -Bdirect patch which now works with glibc 2.5 again! I also removed the hashstyle USE-flag now when both the -Bdirect and --hash-style patches can be applied at the same time (though I don't think they can be used at the same time, please test someone. )
I'm using sys-libs/glibc-2.4.90.20060915.

Now there is sys-libs/glibc-2.5.20061005 from your overlay... In order to update and keep the hash-style patches, I just emerge the new version? Also, should I delete the hashstyle USE-flag for good? Any other change to make on the config files?
You can use 2.5 or the 2.5.20061005 snapshot as hash-style is included in the upstream release now.

Hash-style is included as an option or is it default? Should I delete hashstyle USE-flag or not?

nxsty · Post by **nxsty** » Sat Oct 07, 2006 5:40 pm

JoKo wrote:Hash-style is included as an option or is it default? Should I delete hashstyle USE-flag or not?

It´s on by default now (for all ebuilds).

R-Type · Post by **R-Type** » Sat Oct 07, 2006 7:05 pm

Code: Select all

gcc -DHAVE_CONFIG_H -I. -I/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd -I. -I. -I/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd -I/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/../include -W -Wall -Wstrict-prototypes -Wmissing-prototypes -march=athlon64 -mtune=athlon64 -O3 -pipe -c /var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c  -fPIC -DPIC -o .libs/bfdsort.o
/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c:96: error: expected declaration specifiers or '...' before 'bfd_qsort_closure_func'
/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c:97: warning: no previous prototype for 'bfd_qsort'
/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c: In function 'bfd_qsort':
/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c:128: error: 'cmp' undeclared (first use in this function)
/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c:128: error: (Each undeclared identifier is reported only once
/var/tmp/portage/binutils-2.17.50.0.5/work/binutils-2.17.50.0.5/bfd/bfdsort.c:128: error: for each function it appears in.)
make[4]: *** [bfdsort.lo] Error 1
make[4]: Leaving directory `/var/tmp/portage/binutils-2.17.50.0.5/work/build/bfd'
make[3]: *** [all-recursive] Error 1
make[3]: Leaving directory `/var/tmp/portage/binutils-2.17.50.0.5/work/build/bfd'
make[2]: *** [all] Error 2

I get this when building the new binutils-2.17.50.0.5 build from your overlay. the bdirect useflag is enabled. This is on amd64.

Gergan Penkov · Post by **Gergan Penkov** » Sat Oct 07, 2006 7:23 pm

nxsty does this binutils-bug affect only amd64 or is it creating huge executables on all platforms?

weedy · Post by **weedy** » Sun Oct 08, 2006 12:00 pm

why did we stop filtering flags?

Code: Select all

i586-pc-linux-gnu-gcc -Wl,-O1 -Wl,--enable-new-dtags -Wl,--hash-style=both -Wl,--as-needed -Wl,--sort-common -s  -shared -static-libgcc -Wl,-O1  -Wl,-z,defs -Wl,-dynamic-linker=/lib/ld-linux.so.2  -B/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/csu/  -Wl,--version-script=/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/libc.map -Wl,-soname=libc.so.6 -Wl,-z,combreloc -Wl,-z,relro -Wl,--hash-style=both -nostdlib -nostartfiles -e __libc_main -Wl,-z,now -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/math -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/elf -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/dlfcn -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/nss -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/nis -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/rt -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/resolv -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/crypt -L/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/nptl -Wl,-rpath-link=/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/math:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/elf:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/dlfcn:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/nss:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/nis:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/rt:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/resolv:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/crypt:/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/nptl -o /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/libc.so -T /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/shlib.lds /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/csu/abi-note.o /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/elf/soinit.os /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/libc_pic.os /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/elf/sofini.os /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/elf/interp.os /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/elf/ld.so /var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/debug/stack_chk_fail_local.oS -lgcc
/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/debug/stack_chk_fail_local.oS: In function `__stack_chk_fail_local':
stack_chk_fail_local.c:(.text+0x0): multiple definition of `__stack_chk_fail_local'
/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/libc_pic.os:(.text+0xd0d20): first defined here
collect2: ld returned 1 exit status
make[1]: *** [/var/tmp/portage/glibc-2.5.20061005/work/build-default-i586-pc-linux-gnu/libc.so] Error 1
make[1]: Leaving directory `/var/tmp/portage/glibc-2.5.20061005/work/glibc-2.5'
make: *** [all] Error 2

!!! ERROR: sys-libs/glibc-2.5.20061005 failed.
Call stack:
  ebuild.sh, line 1546:   Called dyn_compile
  ebuild.sh, line 937:   Called src_compile
  glibc-2.5.20061005.ebuild, line 1045:   Called toolchain-glibc_src_compile
  glibc-2.5.20061005.ebuild, line 249:   Called die

nxsty · Post by **nxsty** » Sun Oct 08, 2006 2:08 pm

Gergan Penkov wrote:nxsty does this binutils-bug affect only amd64 or is it creating huge executables on all platforms?

I think it´s mainly an amd64 problem but the bug is in the generic code AFAICT.