OSDN Git Service
Eric Andersen [Sat, 8 Nov 2003 20:41:24 +0000 (20:41 -0000)]
Sigh. I can't type this morning.
Eric Andersen [Sat, 8 Nov 2003 20:33:49 +0000 (20:33 -0000)]
Joakim Tjernlund writes:
Oops, found another ppc 8xx bug.
8xx CPUs may need this as well to work:
Eric Andersen [Sat, 8 Nov 2003 08:58:07 +0000 (08:58 -0000)]
hide the wrapper in with the debian stuff
Eric Andersen [Sat, 8 Nov 2003 08:48:34 +0000 (08:48 -0000)]
Update docs in preparation for release
Eric Andersen [Sat, 8 Nov 2003 08:09:02 +0000 (08:09 -0000)]
kill mention of the wrapper
Eric Andersen [Sat, 8 Nov 2003 08:06:38 +0000 (08:06 -0000)]
For now, disable the iconv utility
Eric Andersen [Sat, 8 Nov 2003 07:42:34 +0000 (07:42 -0000)]
Joakim Tjernlund writes:
> Very interesting. Do you have any suggestions for how
> we could fix our powerpc shared library loader
Removing those instr. comes with a very big performance
penalty. To flush the dcache you will have read up to 8KB
dummy data and to invalidate the icache you will have to
execute up to 16KB nops. I don't know of any other way from
user space.
hmm, actually I think it will work reliable to perform a
store to the same page(s) as the dcbst/icbi will act on. That
way you will make the DTLB Error happen(if any) prior to the
dcbst/icbi. The worst thing that can happen then is a regular
DTLB Miss and that works for dcbst/icbi.
You will have to lookout for if dcbst/icbi crosses a page
boundary. Then you will have to perform a store to both
pages.
Jocke
# And again later writes:
Hi again
I think I know what the problem is. The
PPC_DCBST;PPC_SYNC;PPC_ICBI;PPC_ISYNC sequence is executed
even if no modification has been done i some cases:
_dl_linux_resolver(), the last else has no store for insns[0].
these is a insns[1] = OPCODE_B(delta - 4) that
does not have a PPC_DCBST.
_dl_do_lazy_reloc(), for R_PPC_NONE there is no store.
for R_PPC_JMP_SLOT there is a
insns[1] = OPCODE_B(delta)that does not
have a PPC_DCBST.
_dl_do_reloc(), for R_PPC_COPY there is no store.
for R_PPC_JMP_SLOT there is a
reloc_addr[1] = OPCODE_B(delta) that does not
have a PPC_DCBST.
_dl_init_got(), I THINK that the
PPC_DCBST(plt);
PPC_DCBST(plt+4);
PPC_DCBST(plt+8);
PPC_SYNC;
PPC_ICBI(plt);
PPC_ICBI(plt+4);
PPC_ICBI(plt+8);
PPC_ISYNC;
is off a bit. The address range does not match the sum
of the plt[] and tramp[] address range.
Jocke
# And then later added the comment:
I think that the tramp[] part should be included in the
PPC_DCBST/PPC_ICBI sequence. Then you have to add entries for
plt+12 and plt+16. If the tramp[] part should be excluded,
then all is well.
Jocke
Eric Andersen [Sat, 8 Nov 2003 06:33:36 +0000 (06:33 -0000)]
Fix up several errors related to filename length and errno that
showed up while running the latest LTP testsuite.
-Erik
Eric Andersen [Sat, 8 Nov 2003 00:33:02 +0000 (00:33 -0000)]
Add some basic cvs usage info
Eric Andersen [Fri, 7 Nov 2003 06:22:12 +0000 (06:22 -0000)]
Sigh. I forgot to define INSTALL which, understandable,
broke the install target
Manuel Novoa III [Fri, 7 Nov 2003 06:08:27 +0000 (06:08 -0000)]
Add clock_gettime() submitted by Justus Pendleton <uc@ryoohki.net>.
Eric Andersen [Fri, 7 Nov 2003 05:44:10 +0000 (05:44 -0000)]
remove some junk I added while debugging
Eric Andersen [Fri, 7 Nov 2003 04:13:30 +0000 (04:13 -0000)]
It is remotely possible the utils might even compile this time
around...
Eric Andersen [Thu, 6 Nov 2003 22:38:49 +0000 (22:38 -0000)]
Begin converting the client utils
Eric Andersen [Thu, 6 Nov 2003 22:09:38 +0000 (22:09 -0000)]
Revert change 1.26, which breaks everything. :-(
Manuel Novoa III [Thu, 6 Nov 2003 22:01:03 +0000 (22:01 -0000)]
Sigh... I wasn't checking if the gid field in a group line actually
contained a digit. Also adjust a comment.
Eric Andersen [Thu, 6 Nov 2003 21:23:34 +0000 (21:23 -0000)]
s,(RUNTIME_PREFIX)/,(RUNTIME_PREFIX),g
Eric Andersen [Thu, 6 Nov 2003 21:22:21 +0000 (21:22 -0000)]
s,(DEVEL_PREFIX)/,(DEVEL_PREFIX),g
Manuel Novoa III [Thu, 6 Nov 2003 21:14:22 +0000 (21:14 -0000)]
Add a utility macro.
Manuel Novoa III [Thu, 6 Nov 2003 21:13:03 +0000 (21:13 -0000)]
Oops... don't need recursive mutexes.
Manuel Novoa III [Thu, 6 Nov 2003 20:56:04 +0000 (20:56 -0000)]
Rewrite the pwd.h, grp.h, and shadow.h functions (except lckpwdf/ulckpwdf).
Eric Andersen [Thu, 6 Nov 2003 20:29:35 +0000 (20:29 -0000)]
Add in aliases provided by the non arch specific implementations
Eric Andersen [Thu, 6 Nov 2003 19:44:47 +0000 (19:44 -0000)]
George Thanos writes:
Dear Erik,
We downloded uClibc lattest version from the CVS. Still there are some
minor problems with extra/Configs/Config.e1
You have actually set ARCH_HAS_C_SYMBOL_PREFIX to NO which is not
correct for our architecture. Please apply the patch that will fix the
problem.
Best Regards,
- George
P.S. Patch also removes some irritating comments we have added in the past.
Eric Andersen [Thu, 6 Nov 2003 13:03:42 +0000 (13:03 -0000)]
A few debugging cleanups
Eric Andersen [Thu, 6 Nov 2003 09:39:45 +0000 (09:39 -0000)]
The curse of cut-n-paste
Eric Andersen [Thu, 6 Nov 2003 09:14:26 +0000 (09:14 -0000)]
Add some initial x86 string optimizations. These make no attempt to use nifty
things like mmx/3dnow/etc. These are not inline, and will therefore not be as
fast as modifying the headers to use inlines (and cannot therefore do tricky
things when dealing with const memory). But they should (I hope!) be faster
than their generic equivalents....
More importantly, these should provide a good example for others to follow when
adding arch specific optimizations.
-Erik
Manuel Novoa III [Wed, 5 Nov 2003 19:12:35 +0000 (19:12 -0000)]
Remove erroneous trailing semicolon.
Eric Andersen [Wed, 5 Nov 2003 10:47:01 +0000 (10:47 -0000)]
Fix a misnamed array
Eric Andersen [Wed, 5 Nov 2003 10:46:42 +0000 (10:46 -0000)]
tiny cleanup
Eric Andersen [Wed, 5 Nov 2003 10:12:08 +0000 (10:12 -0000)]
Rework the powerpc lib loader to better match the other architectures
Eric Andersen [Wed, 5 Nov 2003 07:08:20 +0000 (07:08 -0000)]
Rip the guts out of the dynamically generated sysconf.c, and live with a simple
static version. This will need further work later on, but should do the job for
the time being,
Eric Andersen [Wed, 5 Nov 2003 06:44:22 +0000 (06:44 -0000)]
last but certinaly not least, kill off initfini.c
Eric Andersen [Wed, 5 Nov 2003 06:42:49 +0000 (06:42 -0000)]
Add quick hacks to create fake crti.o and crtn.o files. These architectures
need to be fixed properly. I tried, but I was unable to build a cross
toolchain for each of these (using stock binutils and gcc) so it is someone
else's problem to fix them now.
Eric Andersen [Wed, 5 Nov 2003 06:31:40 +0000 (06:31 -0000)]
Kill initfini.awk
Eric Andersen [Wed, 5 Nov 2003 05:57:19 +0000 (05:57 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 05:43:31 +0000 (05:43 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 05:30:07 +0000 (05:30 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 05:20:36 +0000 (05:20 -0000)]
v850 appers to be little endian only
Eric Andersen [Wed, 5 Nov 2003 05:13:53 +0000 (05:13 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 05:08:30 +0000 (05:08 -0000)]
duh. the alpha has an fpu
Eric Andersen [Wed, 5 Nov 2003 05:04:23 +0000 (05:04 -0000)]
Yet more messing with selecting sane default configs
Eric Andersen [Wed, 5 Nov 2003 04:27:32 +0000 (04:27 -0000)]
Yet more config system updates
Eric Andersen [Wed, 5 Nov 2003 03:32:14 +0000 (03:32 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 02:45:34 +0000 (02:45 -0000)]
Fix a buglet that shows up on powerpc
Eric Andersen [Wed, 5 Nov 2003 02:41:22 +0000 (02:41 -0000)]
m68k is always big endian afaik
Eric Andersen [Wed, 5 Nov 2003 02:26:15 +0000 (02:26 -0000)]
powerpc is always big endian
Eric Andersen [Wed, 5 Nov 2003 02:21:28 +0000 (02:21 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 02:16:29 +0000 (02:16 -0000)]
Patch from Peter S. Mazinger to hunt for awk in additional
places....
Eric Andersen [Wed, 5 Nov 2003 02:12:56 +0000 (02:12 -0000)]
Arthur Shipkowski, art ! videon-central ! com, writes:
I've noticed a few people have posted over the last year about problems
compiling programs that use vfork when pthreads are involved. Some
detective work turned up that ptfork.c aliases vfork to fork and then tries
to call the original fork as __libc_fork. This patch removes the aliasing
when there is no MMU present, and uses the same call semantics to call
__libc_vfork. I then added a symbol to the m68k vfork.S to allow vfork to
be called as __libc_vfork.
The same bug exists in the uClibc CVS, and with a possible tweak this patch
should go through there as well.
Obviously, all other platforms need __libc_vfork as a workable means to call
vfork in order for this to work for them.
Let me know if there are any problems with this patch.
Art Shipkowski
Videon Central Software Engineer
(814)235-1111 x307
Eric Andersen [Wed, 5 Nov 2003 02:09:25 +0000 (02:09 -0000)]
Force cris to be little endian. Afaik, there is no big
endian cris architecture.
Eric Andersen [Wed, 5 Nov 2003 02:04:16 +0000 (02:04 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 01:54:45 +0000 (01:54 -0000)]
Oops
Eric Andersen [Wed, 5 Nov 2003 01:44:16 +0000 (01:44 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 01:14:46 +0000 (01:14 -0000)]
Do not provide a default for endianness. This needs to be selected
on a per-arch basis, or left to the user to choose.
Eric Andersen [Wed, 5 Nov 2003 01:08:25 +0000 (01:08 -0000)]
Kill off "mipsel" and just use "mips"
Eric Andersen [Wed, 5 Nov 2003 00:43:09 +0000 (00:43 -0000)]
Continue the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Wed, 5 Nov 2003 00:29:20 +0000 (00:29 -0000)]
Begin the conversion to using per-arch crti.S and crtn.S
Eric Andersen [Tue, 4 Nov 2003 20:58:12 +0000 (20:58 -0000)]
As Peter S. Mazinger has pointed out, the config system wants
some defaults. So give it some empty defaults and let people
select their own options.
Eric Andersen [Tue, 4 Nov 2003 20:46:53 +0000 (20:46 -0000)]
Peter S. Mazinger writes:
Hello!
The latest changes document ldd in RUNTIME_PREFIX/bin, but it is installed
in RUNTIME_PREFIX/usr/bin
Peter
Eric Andersen [Tue, 4 Nov 2003 13:35:44 +0000 (13:35 -0000)]
Naming things this way will be much easier to deal with
Eric Andersen [Tue, 4 Nov 2003 13:15:05 +0000 (13:15 -0000)]
Kill off the gcc wrapper. It has served us well, but there comes a time when
90% of correct is simply not good enough. Some people will not be very happy
about the decision to kill the wrapper toolchain. Sorry, but a real toolchain
is the One True Way(tm).
-Erik
Eric Andersen [Tue, 4 Nov 2003 13:07:50 +0000 (13:07 -0000)]
Rework the config system. Better utilize the Kconfig language
which should simplify enabling arbitrary architectures.
-Erik
Eric Andersen [Tue, 4 Nov 2003 09:00:56 +0000 (09:00 -0000)]
minor cleanup
Eric Andersen [Tue, 4 Nov 2003 08:51:17 +0000 (08:51 -0000)]
Minor cleanup
Eric Andersen [Tue, 4 Nov 2003 05:49:28 +0000 (05:49 -0000)]
Minor makefile tweaks
Eric Andersen [Mon, 3 Nov 2003 09:50:25 +0000 (09:50 -0000)]
Begin preparations for the next release
Eric Andersen [Mon, 3 Nov 2003 09:48:01 +0000 (09:48 -0000)]
Remove warning that apparently is not needed for e1
Tobias Anderberg [Mon, 3 Nov 2003 08:43:01 +0000 (08:43 -0000)]
Updated to match the other architectures. Fixes problem with "fcntl:
function not implemented".
Eric Andersen [Sun, 2 Nov 2003 21:55:06 +0000 (21:55 -0000)]
Additional passwd and group tests
Eric Andersen [Sun, 2 Nov 2003 21:35:27 +0000 (21:35 -0000)]
Implement getgrent_r. Rework getpwent and getgrent a bit further
Eric Andersen [Sun, 2 Nov 2003 10:03:23 +0000 (10:03 -0000)]
Both setegid and seteuid were implemented suboptimally, such that
we were unable to switch back to the original saved group/user ID.
-Erik
Eric Andersen [Sun, 2 Nov 2003 10:01:11 +0000 (10:01 -0000)]
Add some missing prototypes
Eric Andersen [Sun, 2 Nov 2003 10:00:45 +0000 (10:00 -0000)]
Be extra careful to check uid and gid converstions to kernel types
Eric Andersen [Sun, 2 Nov 2003 09:45:14 +0000 (09:45 -0000)]
Make the syscall locally, avoid an extern
Manuel Novoa III [Sun, 2 Nov 2003 08:29:10 +0000 (08:29 -0000)]
If realloc failed, we'd lose the pointer to the exit function table.
Manuel Novoa III [Sun, 2 Nov 2003 07:39:27 +0000 (07:39 -0000)]
Hopefully fix the struct tm extension problem once and for all.
Also fix a dst-related bug which caused the use of uninitialized data.
Eric Andersen [Sat, 1 Nov 2003 04:49:27 +0000 (04:49 -0000)]
Add some additional pwd/grp tests, to prevent me from
breaking obvious things in the future.
-Erik
Eric Andersen [Sat, 1 Nov 2003 04:40:10 +0000 (04:40 -0000)]
Fix things (properly) to open /etc/passd and /etc/group if
they have not yet been opened.
My last try was completely and embarrasingly broken.
-Erik
Eric Andersen [Sat, 1 Nov 2003 03:55:29 +0000 (03:55 -0000)]
this script is no longer used, and can be removed.
Manuel Novoa III [Fri, 31 Oct 2003 23:50:25 +0000 (23:50 -0000)]
Fix Erik's typo.
Manuel Novoa III [Fri, 31 Oct 2003 23:26:03 +0000 (23:26 -0000)]
Some more soft float fixes... for arm in particular (libfloat).
Remove the ADD_LIBGCC_FUNCTIONS option and do things the right way.
Either we have a shared libgcc available, or the libgcc routines
aren't PIC and don't belong in the shared libc anyway.
Eric Andersen [Fri, 31 Oct 2003 20:17:01 +0000 (20:17 -0000)]
getpwent(), getpwent_r(), and getgrent(), getgrent_r() would all
fail if you had not previously called setpwent() or setgrent()
respectively. Oops. My bad.
-Erik
Eric Andersen [Fri, 31 Oct 2003 06:19:53 +0000 (06:19 -0000)]
gdb wants to dlopen "libthread_db.so.1", so be sure to assign
the correct soname
Eric Andersen [Sat, 25 Oct 2003 06:35:29 +0000 (06:35 -0000)]
properly deal with soft-float when profiling as well
Manuel Novoa III [Sat, 25 Oct 2003 02:59:33 +0000 (02:59 -0000)]
Add -msoft-float to SAFECFLAGS when necessary.
Eric Andersen [Fri, 24 Oct 2003 04:52:29 +0000 (04:52 -0000)]
Sigh. It seems some stupid programs expect this...
Eric Andersen [Fri, 24 Oct 2003 04:12:42 +0000 (04:12 -0000)]
Disable optimizations with debugging
Eric Andersen [Fri, 24 Oct 2003 04:11:50 +0000 (04:11 -0000)]
better tuning for xscale that copes with generally available toolchains
Eric Andersen [Thu, 23 Oct 2003 07:10:50 +0000 (07:10 -0000)]
Peter Kjellerstedt writes:
The attached patch performs a clean up of
extra/gcc-uClibc/gcc-uClibc.c:
* Use NULL for pointers and '\0' for characters;
not the opposite...
* Fixed an (assumed) segv if --uclibc-cc was used.
* Skip the argument to -x so it isn't considered
to be a source file.
* Simplified the loop in xstrcat().
* A little white space clean up.
//Peter
Eric Andersen [Wed, 22 Oct 2003 19:43:08 +0000 (19:43 -0000)]
Peter S. Mazinger pointed out that I missed a spot. I should
stop applying patches by hand...
Eric Andersen [Wed, 22 Oct 2003 09:52:24 +0000 (09:52 -0000)]
Per suggestion and patch from Ken Staton, emulates poll using
select for older 2.0 kernels where poll is missing.
Eric Andersen [Wed, 22 Oct 2003 09:29:40 +0000 (09:29 -0000)]
Atsushi Nemoto writes:
uClibc's libm has isinf/isinff, isnan/isnanf but not finite/finitef.
Here is a patch.
Eric Andersen [Wed, 22 Oct 2003 09:15:57 +0000 (09:15 -0000)]
It seems we need to also define SI_LOAD_SHIFT since that was
also exported by the Linux kernel.h header that we do not include.
Eric Andersen [Tue, 21 Oct 2003 06:49:01 +0000 (06:49 -0000)]
Those crazy binutils folks changed their abi. Add this so older
apps can cope gracefully.
Eric Andersen [Tue, 21 Oct 2003 00:52:14 +0000 (00:52 -0000)]
Doh! In include/bits/uClibc_config.h when we define things, we appropriately
prepend "__". Unfortunately, when we #undef things, we do not prepend the
"__". This results in collateral damage to unsuspecting applications as we
undefine random unrelated macros. Oops.
For example, when compiling xfree86, libGLcore.a defines MALLOC for its own usage.
But include/bits/uClibc_config.h then does an
#undef MALLOC
#define __MALLOC_930716__ 1
which inadvertantly trashes the define from libGLcore.... Ouch.
The autogenerated include/bits/uClibc_config.h should instead have
contained:
#undef __MALLOC__
#define __MALLOC_930716__ 1
This patch makes that happen.
Eric Andersen [Mon, 20 Oct 2003 20:13:19 +0000 (20:13 -0000)]
Fix a stupid bug that caused uClibc to never provide the correct
fpu_control.h header file, since the correct arch specific one was
always later overwritten by the generic one. oops.
-Erik
Eric Andersen [Mon, 20 Oct 2003 18:23:28 +0000 (18:23 -0000)]
Patch from Peter S. Mazinger:
patch to allow building uClibc with busybox's chmod
Eric Andersen [Mon, 20 Oct 2003 06:17:00 +0000 (06:17 -0000)]
Merge some newer/updated bits from the glibc elf.h
Eric Andersen [Sun, 19 Oct 2003 00:28:47 +0000 (00:28 -0000)]
We were failing to properly set h_errno on success, which could
cause gethostbyaddr_r to keep looping allocating more and more
memory each time till alloca finally caused a segfault. Ugh.
This fixes that as well...
-Erik
Eric Andersen [Sat, 18 Oct 2003 12:43:31 +0000 (12:43 -0000)]
more install changes