OSDN Git Service

uclinux-h8/uClibc.git
17 years agouse __FDPIC__ from the compiler rather than setting our own
Mike Frysinger [Thu, 8 Feb 2007 21:55:28 +0000 (21:55 -0000)]
use __FDPIC__ from the compiler rather than setting our own

17 years agorevert 17821, wait for better test cases
Eric Andersen [Thu, 8 Feb 2007 21:07:26 +0000 (21:07 -0000)]
revert 17821, wait for better test cases

17 years agodo not use ghbn_buf after it goes out of scope
Eric Andersen [Thu, 8 Feb 2007 21:02:26 +0000 (21:02 -0000)]
do not use ghbn_buf after it goes out of scope

17 years agoimport a slightly modified realpath() test case from glibc
Mike Frysinger [Thu, 8 Feb 2007 19:07:11 +0000 (19:07 -0000)]
import a slightly modified realpath() test case from glibc

17 years agoMiguel Angel Alvarez writes:
Eric Andersen [Thu, 8 Feb 2007 18:38:50 +0000 (18:38 -0000)]
Miguel Angel Alvarez writes:

Hi. I have found a bug in realpath that did not allow me to use it with
symlinks. Although I am using 0.9.28.1, the bug is present in 0.9.28 and
also in the trunk.

I attach a suggested patch. Please let me know if the solution is correct.

17 years agohandle file_name==NULL and file_name=="" as required by POSIX
Mike Frysinger [Thu, 8 Feb 2007 18:25:57 +0000 (18:25 -0000)]
handle file_name==NULL and file_name=="" as required by POSIX

17 years agorevert previous change; reading POSIX spec some more says we need to set errno to...
Mike Frysinger [Thu, 8 Feb 2007 18:24:46 +0000 (18:24 -0000)]
revert previous change; reading POSIX spec some more says we need to set errno to EINVAL if filename is NULL

17 years agowe crash if the first argument to realpath() is non-null as well so lets mark both...
Mike Frysinger [Thu, 8 Feb 2007 18:21:14 +0000 (18:21 -0000)]
we crash if the first argument to realpath() is non-null as well so lets mark both arguments as being non-NULL

17 years agoPaul Brook writes:
Mike Frysinger [Thu, 8 Feb 2007 18:05:44 +0000 (18:05 -0000)]
Paul Brook writes:
I'm seeing a hang in applications that open and close files with a
non-threaded uClibc.  There is code in fclose() to remove a file handle from
_stdio_openlist however it's commented out, allegedly because it is dead.
_stdio_openlist is used in several places, including _stdio_term when a
program exits. fclose() frees the struct, so I don't see how the code can
possibly be redundant.

In the __UCLIBC_HAS_THREADS__ && __STDIO_BUFFERS case the file is removed from
_stdio_openlist by __STDIO_OPENLIST_DEC_USE.

The attached patch fixes this bug be re-enabling the code.

17 years agopull in features.h so we have access to config defines
Mike Frysinger [Thu, 8 Feb 2007 16:49:07 +0000 (16:49 -0000)]
pull in features.h so we have access to config defines

17 years agosh crti.S needs features.h to resolve __HAVE_SHARED__, trivial patch
Paul Mundt [Wed, 7 Feb 2007 11:02:37 +0000 (11:02 -0000)]
sh crti.S needs features.h to resolve __HAVE_SHARED__, trivial patch
from Carmelo Amoroso <carmelo73@gmail.com>.

17 years agono need to duplicate WCHAR check
Mike Frysinger [Mon, 5 Feb 2007 16:16:51 +0000 (16:16 -0000)]
no need to duplicate WCHAR check

17 years ago- silence warning about using an undefined token
Bernhard Reutner-Fischer [Mon, 5 Feb 2007 12:33:30 +0000 (12:33 -0000)]
- silence warning about using an undefined token

17 years ago- silence warning about using an undefined token
Bernhard Reutner-Fischer [Mon, 5 Feb 2007 12:17:12 +0000 (12:17 -0000)]
- silence warning about using an undefined token

17 years agonew cheesy test by Denis Vlasenko to trigger fclose loop
Mike Frysinger [Mon, 5 Feb 2007 01:17:09 +0000 (01:17 -0000)]
new cheesy test by Denis Vlasenko to trigger fclose loop

17 years agoAdd support for PowerPC e500, libc part only.
Joakim Tjernlund [Sun, 4 Feb 2007 21:25:42 +0000 (21:25 -0000)]
Add support for PowerPC e500, libc part only.
Math support to be done.
From Steve Papacharalambous, stevep@freescale.com

17 years agofixup includes so we compile again
Mike Frysinger [Sun, 4 Feb 2007 09:43:11 +0000 (09:43 -0000)]
fixup includes so we compile again

17 years agothe sun rpc code has too many alias violations to worry over now; just disable strict...
Mike Frysinger [Sun, 4 Feb 2007 09:42:38 +0000 (09:42 -0000)]
the sun rpc code has too many alias violations to worry over now; just disable strict-aliasing until someone gets the balls to clean this stuff up

17 years agofix compiling and touchup style
Mike Frysinger [Sun, 4 Feb 2007 09:34:13 +0000 (09:34 -0000)]
fix compiling and touchup style

17 years agosync some stuff from glibc to fix test build failure
Mike Frysinger [Sun, 4 Feb 2007 09:24:56 +0000 (09:24 -0000)]
sync some stuff from glibc to fix test build failure

17 years agoMake sure struct stat is identical to struct stat64 if
Peter Kjellerstedt [Fri, 2 Feb 2007 12:32:27 +0000 (12:32 -0000)]
Make sure struct stat is identical to struct stat64 if
__USE_FILE_OFFSET64 is defined. This was the case for all
architectures except those that use packed structures by default
(i.e., CRIS).

17 years agouse sane line breaks
Eric Andersen [Fri, 2 Feb 2007 02:20:59 +0000 (02:20 -0000)]
use sane line breaks

17 years agoreplace bits/syscalls.h with a comment ... an empty file makes people think this...
Mike Frysinger [Fri, 2 Feb 2007 01:42:51 +0000 (01:42 -0000)]
replace bits/syscalls.h with a comment ... an empty file makes people think this is a bug rather than by design

17 years agoRonald Maeder writes:
Eric Andersen [Fri, 2 Feb 2007 01:29:10 +0000 (01:29 -0000)]
Ronald Maeder writes:
I have successfully made gethostbyname_r(), res_init(), and gethostid() fully
reentrant.  In addition, I have added a NULL check to inet_aton().  This is
where SEG FAULTs were coming from when gethostbyname_r() was called.

17 years agouse the correct lock for access to __nameserver[]
Eric Andersen [Fri, 2 Feb 2007 01:01:51 +0000 (01:01 -0000)]
use the correct lock for access to __nameserver[]

17 years agofix lock naming
Eric Andersen [Fri, 2 Feb 2007 00:37:06 +0000 (00:37 -0000)]
fix lock naming

17 years agoRonald Maeder writes:
Eric Andersen [Fri, 2 Feb 2007 00:30:30 +0000 (00:30 -0000)]
Ronald Maeder writes:

Hi Erik,

Thanks for all your great work.  I found a set of bugs in
resolv.c .  Basically, there is code that looks like:

        BIGLOCK;
        __nameserversXX=__nameservers;
        __nameserverXX=__nameserver;
        BIGUNLOCK;
        i = __dns_lookup(dname, type, __nameserversXX, __nameserverXX, &packet, &a);

which is a problem because the declarations are

int __nameservers;
char * __nameserver[MAX_SERVERS];
int __searchdomains;
char * __searchdomain[MAX_SEARCH];

so you can see that __nameserver is a pointer.  Copying the
pointer to __nameserverXX doesn't protect the global variable
space.  I have attached a patch and the new file.  I hope you
will incorporate these bug fixes.  I spent quite a bit of time
tracking them down.

Many thanks,

Ron

17 years agofix obvious bug in ipv4/ipv6 resolving. When not using AF_INET,
Eric Andersen [Fri, 2 Feb 2007 00:23:55 +0000 (00:23 -0000)]
fix obvious bug in ipv4/ipv6 resolving.  When not using AF_INET,
gethostbyname2_r tries to resolve an ipv6 address from /etc/hosts using
get_hosts_byname_r, but with AF_INET instead of the supplied address family.
This returns ipv4 addresses marked as ipv6 ones.  Fix from nbd.

17 years agodrive home the point people should not be using _syscall#() macros
Mike Frysinger [Thu, 1 Feb 2007 07:30:37 +0000 (07:30 -0000)]
drive home the point people should not be using _syscall#() macros

17 years agodocument future need for larger time values
Mike Frysinger [Thu, 1 Feb 2007 07:16:04 +0000 (07:16 -0000)]
document future need for larger time values

17 years ago- should look if LFS is enabled as suggested by pkj in #1131. This generates identica...
Bernhard Reutner-Fischer [Tue, 30 Jan 2007 17:26:38 +0000 (17:26 -0000)]
- should look if LFS is enabled as suggested by pkj in #1131. This generates identical cc1 input for LFS enabled (compared to before r17652) but makes sure that we pickup the non-LFS if LFS is turned off (like r17652 did in this case).

17 years ago- fix compilation of GNU glob with !LFS
Bernhard Reutner-Fischer [Mon, 29 Jan 2007 22:40:38 +0000 (22:40 -0000)]
- fix compilation of GNU glob with !LFS
  Sounds like a typo, really

17 years agoRichard Sandiford writes:
Mike Frysinger [Mon, 29 Jan 2007 17:52:29 +0000 (17:52 -0000)]
Richard Sandiford writes:
However, retesting on m68k showed up a problem that had appeared in
uClibc since the last time I tried.  Specifically, revision 15785 did:

  -#define HEAP_GRANULARITY       (sizeof (HEAP_GRANULARITY_TYPE))
  +#define HEAP_GRANULARITY       (__alignof__ (HEAP_GRANULARITY_TYPE))

  -#define MALLOC_ALIGNMENT       (sizeof (double))
  +#define MALLOC_ALIGNMENT       (__alignof__ (double))

The problem is that

  (a) MALLOC_HEADER_SIZE == MALLOC_ALIGNMENT
  (b) the header contains a size value of type size_t
  (c) sizeof (size_t) is 4 on m68k, but...
  (d) __alignof__ (double) is only 2 (the largest alignment used on m68k)

So we only allocate 2 bytes for the 4-byte header, and the least
significant 2 bytes of the size are in the user's area rather than
the header.  The patch below fixes that problem by redefining
MALLOC_HEADER_SIZE to:

  MAX (MALLOC_ALIGNMENT, sizeof (size_t))

(but without the help of the MAX macro ;)).  However, we really would
like to have word alignment on Coldfire.  It makes a big performance
difference, and because we have to allocate a 4-byte header anyway,
what wastage there is will be confined to the end of the allocated block.
Any wastage will also be limited to 2 bytes per allocation compared to
the current alignment.

I've therefore used the __aligned__ type attribute to create a double
type that has at least sizeof (size_t) bytes of alignment.  I've
introduced a new __attribute_aligned__ macro for this.  It might seem
silly protecting against old or non-GNU compilers here, but the extra
alignment is only an optimisation, and having the macro is more in the
spirit of the other attribute code.

17 years ago- s/if USE___THREAD/if defined USE___THREAD && USE___THREAD/g
Bernhard Reutner-Fischer [Mon, 29 Jan 2007 14:56:56 +0000 (14:56 -0000)]
- s/if USE___THREAD/if defined USE___THREAD && USE___THREAD/g

17 years ago- see if defined foo before trying to look at the value of foo (that should be 0...
Bernhard Reutner-Fischer [Mon, 29 Jan 2007 14:33:27 +0000 (14:33 -0000)]
- see if defined foo before trying to look at the value of foo (that should be 0 in this case. Avoids some warnings).

17 years ago- sync with glibc (and provide a check for icc).
Bernhard Reutner-Fischer [Mon, 29 Jan 2007 14:31:16 +0000 (14:31 -0000)]
- sync with glibc (and provide a check for icc).
  Only affects gcc <2.7 or non-gcc compilers that bailed earlier due to missing int64_t

17 years agopull in unistd.h for syscall() prototype
Mike Frysinger [Mon, 29 Jan 2007 04:32:20 +0000 (04:32 -0000)]
pull in unistd.h for syscall() prototype

17 years agouse ifdef rather than if for stack direction
Mike Frysinger [Mon, 29 Jan 2007 04:28:46 +0000 (04:28 -0000)]
use ifdef rather than if for stack direction

17 years agofix from Atsushi Nemoto for displaying of 64bit types
Mike Frysinger [Mon, 29 Jan 2007 04:07:37 +0000 (04:07 -0000)]
fix from Atsushi Nemoto for displaying of 64bit types

17 years agofix from Roman Kononov for reading in of 64bit types
Mike Frysinger [Mon, 29 Jan 2007 04:04:19 +0000 (04:04 -0000)]
fix from Roman Kononov for reading in of 64bit types

17 years agomake sure reading 64bit types works
Mike Frysinger [Mon, 29 Jan 2007 04:03:31 +0000 (04:03 -0000)]
make sure reading 64bit types works

17 years agomake sure 64bit output is correct
Mike Frysinger [Mon, 29 Jan 2007 03:59:09 +0000 (03:59 -0000)]
make sure 64bit output is correct

17 years agosince these functions get exported, namespace the arguments
Mike Frysinger [Mon, 29 Jan 2007 03:22:36 +0000 (03:22 -0000)]
since these functions get exported, namespace the arguments

17 years agomark local function static
Mike Frysinger [Mon, 29 Jan 2007 03:21:52 +0000 (03:21 -0000)]
mark local function static

17 years agofixup shadowed warnings
Mike Frysinger [Mon, 29 Jan 2007 03:18:59 +0000 (03:18 -0000)]
fixup shadowed warnings

17 years agomark some funcs as noreturn
Mike Frysinger [Mon, 29 Jan 2007 03:06:31 +0000 (03:06 -0000)]
mark some funcs as noreturn

17 years agofixup prototype warnings
Mike Frysinger [Mon, 29 Jan 2007 02:56:45 +0000 (02:56 -0000)]
fixup prototype warnings

17 years agocheck to see if __USE_FILE_OFFSET64 is defined rather than if it evaluates to true
Mike Frysinger [Mon, 29 Jan 2007 02:34:01 +0000 (02:34 -0000)]
check to see if __USE_FILE_OFFSET64 is defined rather than if it evaluates to true

17 years agomake error output a little more helpful
Mike Frysinger [Mon, 29 Jan 2007 02:31:59 +0000 (02:31 -0000)]
make error output a little more helpful

17 years agofixup warnings
Mike Frysinger [Mon, 29 Jan 2007 02:31:43 +0000 (02:31 -0000)]
fixup warnings

17 years agofix test to work even if CPPFLAGS contains -DNDEBUG
Mike Frysinger [Mon, 29 Jan 2007 02:31:32 +0000 (02:31 -0000)]
fix test to work even if CPPFLAGS contains -DNDEBUG

17 years agosync some fixes from glibc
Mike Frysinger [Mon, 29 Jan 2007 02:04:51 +0000 (02:04 -0000)]
sync some fixes from glibc

17 years agomissed an underscore in previous commit
Mike Frysinger [Mon, 29 Jan 2007 01:43:26 +0000 (01:43 -0000)]
missed an underscore in previous commit

17 years agoBernhard Fischer writes: add undef/nonnull warnings to EXTRA_WARNINGS
Mike Frysinger [Sun, 28 Jan 2007 23:05:17 +0000 (23:05 -0000)]
Bernhard Fischer writes: add undef/nonnull warnings to EXTRA_WARNINGS

17 years agoBernhard Fischer writes: note how uClibc chooses to handle __resolved==NULL
Mike Frysinger [Sun, 28 Jan 2007 23:04:06 +0000 (23:04 -0000)]
Bernhard Fischer writes: note how uClibc chooses to handle __resolved==NULL

17 years agosync with glibc as pointed out by Bernhard Fischer
Mike Frysinger [Sun, 28 Jan 2007 23:02:24 +0000 (23:02 -0000)]
sync with glibc as pointed out by Bernhard Fischer

17 years ago- commentary typo fix
Bernhard Reutner-Fischer [Sun, 28 Jan 2007 13:16:58 +0000 (13:16 -0000)]
- commentary typo fix

17 years agofix from Bernd Schmidt to make sure _dl_debug_state isnt optimized away and does...
Mike Frysinger [Sun, 28 Jan 2007 10:32:59 +0000 (10:32 -0000)]
fix from Bernd Schmidt to make sure _dl_debug_state isnt optimized away and does not generate relocs

17 years agoas pointed out by Jocke, use "2" for actual error messages and _dl_debug_file for...
Mike Frysinger [Sun, 28 Jan 2007 10:29:23 +0000 (10:29 -0000)]
as pointed out by Jocke, use "2" for actual error messages and _dl_debug_file for debug messages

17 years agoinitial waitid() implementation
Mike Frysinger [Sun, 28 Jan 2007 10:23:07 +0000 (10:23 -0000)]
initial waitid() implementation

17 years agoneed string.h for memset()
Mike Frysinger [Sun, 28 Jan 2007 08:20:00 +0000 (08:20 -0000)]
need string.h for memset()

17 years agonote 0.9.28.1 release
Mike Frysinger [Sun, 28 Jan 2007 08:14:03 +0000 (08:14 -0000)]
note 0.9.28.1 release

17 years agoits a new year
Mike Frysinger [Sun, 28 Jan 2007 07:59:59 +0000 (07:59 -0000)]
its a new year

17 years agomake sure we clear out the buffer so that random garbage on the stack doesnt screw...
Mike Frysinger [Sun, 28 Jan 2007 07:26:32 +0000 (07:26 -0000)]
make sure we clear out the buffer so that random garbage on the stack doesnt screw us up

17 years agoPaul Brook writes:
Mike Frysinger [Sun, 28 Jan 2007 06:58:44 +0000 (06:58 -0000)]
Paul Brook writes:
The patch below tweaks the ARM nommu startup code so that doubleword stack
alignment is preserved.  This is required on EABI targets.

17 years agofix typo as pointed out by Tony Wu and cleanup source
Mike Frysinger [Sun, 28 Jan 2007 06:55:56 +0000 (06:55 -0000)]
fix typo as pointed out by Tony Wu and cleanup source

17 years agodont rely on random stack state to trigger this issue; force the default stat structs...
Mike Frysinger [Sun, 28 Jan 2007 06:29:33 +0000 (06:29 -0000)]
dont rely on random stack state to trigger this issue; force the default stat structs to alternative bit masks

17 years agonew test taken from stat issue with new tar
Mike Frysinger [Sun, 28 Jan 2007 06:18:24 +0000 (06:18 -0000)]
new test taken from stat issue with new tar

17 years agouse .bz2 rather than .gz
Mike Frysinger [Sun, 28 Jan 2007 05:37:54 +0000 (05:37 -0000)]
use .bz2 rather than .gz

17 years agofix from blackfin repo: Jie Zhang writes: update __kernel_ipc_pid_t, __kernel_uid_t...
Mike Frysinger [Sat, 27 Jan 2007 08:19:09 +0000 (08:19 -0000)]
fix from blackfin repo: Jie Zhang writes: update __kernel_ipc_pid_t, __kernel_uid_t and __kernel_gid_t types to match actual kernel types

17 years agofix from blackfin repo: Jie Zhang writes: Passing NULL as the second argument to...
Mike Frysinger [Sat, 27 Jan 2007 08:17:38 +0000 (08:17 -0000)]
fix from blackfin repo: Jie Zhang writes: Passing NULL as the second argument to utimes () if the second argument to utime () is NULL as the kernel will take care of getting/setting current time

17 years ago- fix format of comment
Bernhard Reutner-Fischer [Fri, 26 Jan 2007 08:55:28 +0000 (08:55 -0000)]
- fix format of comment

17 years agodont actually have to worry about local changes as svn export takes care of that
Mike Frysinger [Fri, 26 Jan 2007 00:31:29 +0000 (00:31 -0000)]
dont actually have to worry about local changes as svn export takes care of that

17 years agoFix libdl bug reported by Cedric Hombourger in
Joakim Tjernlund [Fri, 26 Jan 2007 00:25:10 +0000 (00:25 -0000)]
Fix libdl bug reported by Cedric Hombourger in
http://uclibc.org/lists/uclibc/2007-January/017165.html

17 years agofix from Bryan Wu: return NULL upon error, not the NULL string
Mike Frysinger [Wed, 24 Jan 2007 10:09:41 +0000 (10:09 -0000)]
fix from Bryan Wu: return NULL upon error, not the NULL string

17 years agoRecognize _SC_MONOTONIC_CLOCK in sysconf() even if __NR_clock_getres
Peter Kjellerstedt [Mon, 22 Jan 2007 09:20:55 +0000 (09:20 -0000)]
Recognize _SC_MONOTONIC_CLOCK in sysconf() even if __NR_clock_getres
is not defined.

17 years ago- repair misplaced #endif that got introduced in r17410. Fixes compilation.
Bernhard Reutner-Fischer [Sun, 21 Jan 2007 10:46:53 +0000 (10:46 -0000)]
- repair misplaced #endif that got introduced in r17410. Fixes compilation.

17 years agoDisable declarations for NPTL.
"Steven J. Hill" [Sat, 20 Jan 2007 22:03:02 +0000 (22:03 -0000)]
Disable declarations for NPTL.

17 years agoThe case for _SC_MONOTONIC_CLOCK should only exist if 'clock_getres' does.
"Steven J. Hill" [Sat, 20 Jan 2007 20:40:53 +0000 (20:40 -0000)]
The case for _SC_MONOTONIC_CLOCK should only exist if 'clock_getres' does.

17 years ago- Trim superfluous trailing whitespace. No obj-code changes.
Bernhard Reutner-Fischer [Sat, 20 Jan 2007 18:32:10 +0000 (18:32 -0000)]
- Trim superfluous trailing whitespace. No obj-code changes.

17 years agoI fixed that item long ago.
Joakim Tjernlund [Thu, 11 Jan 2007 22:52:23 +0000 (22:52 -0000)]
I fixed that item long ago.

17 years agoPeter Mazinger pointed out that my last commit was faulty.
Joakim Tjernlund [Thu, 11 Jan 2007 22:39:13 +0000 (22:39 -0000)]
Peter Mazinger pointed out that my last commit was faulty.
This should fix it.

17 years agomerge from blackfin.uclinux.org: Adjust
Mike Frysinger [Wed, 10 Jan 2007 22:18:39 +0000 (22:18 -0000)]
merge from blackfin.uclinux.org: Adjust
crt1.S so that __uClibc_main is called properly

17 years agoFix SEGV for static builds in exit() path.
Joakim Tjernlund [Wed, 10 Jan 2007 22:03:34 +0000 (22:03 -0000)]
Fix SEGV for static builds in exit() path.
Leave the now obsolete libc/misc/pthread dir in for now.

17 years agoneed to keep __check_rhosts_file exported for proper rhost control
Mike Frysinger [Wed, 10 Jan 2007 17:46:19 +0000 (17:46 -0000)]
need to keep __check_rhosts_file exported for proper rhost control

17 years agosync with upstream blackfin.uclinux.org
Mike Frysinger [Wed, 10 Jan 2007 17:02:16 +0000 (17:02 -0000)]
sync with upstream blackfin.uclinux.org

17 years agoAurelien Jacobs writes:
Joakim Tjernlund [Tue, 9 Jan 2007 22:38:55 +0000 (22:38 -0000)]
Aurelien Jacobs writes:

After our last update a new compilation breakage appeared when compiling
gcc:

/home/aurel/geex/geexbox/build.i386/toolchain/i386-pc-linux-uclibc/sysroot/usr/include/bits/uClibc_stdio.h:346: error: expected initializer before 'attribute_hidden'

Some investigations revealed that the following changeset is responsible
for this breakage:

http://uclibc.org/cgi-bin/viewcvs.cgi/trunk/uClibc/libc/sysdeps/linux/common/bits/uClibc_stdio.h?rev=16801&r1=16793&r2=16801&diff_format=h

Reverting just one hunk of this changeset fixed the problem us.
I'm not sure this is the right fix, but please see attached patch.

Look good to me.

17 years agoJoseph S. Myers writes:
Joakim Tjernlund [Tue, 9 Jan 2007 22:30:16 +0000 (22:30 -0000)]
Joseph S. Myers writes:

ELF symbol names are arbitrary 0-terminated sequences of bytes, and the
ELF hash function is defined in the ELF specification to use unsigned
char.  Thus uClibc's _dl_elf_hash, using plain char, breaks when char is
signed and symbol names contain bytes with the high bit set, as with GCC's
ucnid-* tests.  This patch fixes this problem.

17 years agoSupport SecurePLTs for PowerPC. You need a toolchain that supports
Joakim Tjernlund [Fri, 5 Jan 2007 10:20:37 +0000 (10:20 -0000)]
Support SecurePLTs for PowerPC. You need a toolchain that supports
config option --enable-secureplt. The assembler must also
supports R_PPC_REL16* relocations. gcc 4.1.1 and binutils 2.17
is known to do this.

17 years agoAtle Nissestad writes: The attached patch fixes compilation of the current svn on...
Mike Frysinger [Fri, 5 Jan 2007 09:09:22 +0000 (09:09 -0000)]
Atle Nissestad writes: The attached patch fixes compilation of the current svn on the nios2 platform, and updates the crt1/n/i.S files to get CTOR/DTOR-support to work.

17 years ago- s/Attemt/Attempt/
Bernhard Reutner-Fischer [Wed, 3 Jan 2007 22:53:41 +0000 (22:53 -0000)]
- s/Attemt/Attempt/

17 years ago- r16732 forgot to checkin the sync of the __UCLIBC_SUSV3_LEGACY_MACROS__ from string.h
Bernhard Reutner-Fischer [Sat, 23 Dec 2006 14:12:40 +0000 (14:12 -0000)]
- r16732 forgot to checkin the sync of the __UCLIBC_SUSV3_LEGACY_MACROS__ from string.h

17 years ago- provide a hidden_def for fputc_unlocked that is aliased to __fputc_unlocked.
Bernhard Reutner-Fischer [Fri, 22 Dec 2006 14:03:06 +0000 (14:03 -0000)]
- provide a hidden_def for fputc_unlocked that is aliased to __fputc_unlocked.
  Fixes undefined references to __fputc_unlocked when calling fputc().

17 years agomake sure UCLIBC_LDSO can be found via path
Mike Frysinger [Wed, 20 Dec 2006 16:17:33 +0000 (16:17 -0000)]
make sure UCLIBC_LDSO can be found via path

17 years agoPaul Brook writes:
Mike Frysinger [Wed, 20 Dec 2006 04:30:25 +0000 (04:30 -0000)]
Paul Brook writes:
When building an arm-uclinux toolchain withthreading disabled I get:

libc/stdio/getchar.c:26: error: 'getchar' aliased to undefined
symbol 'getchar_unlocked'

It looks like it's missing libc_hidden_def(getchar_unlocked). Patch below
fixes this.

17 years agorename local "brk" var so it doesnt namespace collide with the brk() function
Mike Frysinger [Sun, 17 Dec 2006 20:56:01 +0000 (20:56 -0000)]
rename local "brk" var so it doesnt namespace collide with the brk() function

17 years agofor now, dont allow mmu users to select file format
Mike Frysinger [Sun, 17 Dec 2006 20:49:10 +0000 (20:49 -0000)]
for now, dont allow mmu users to select file format

17 years agoactually scrub it, dont just comment it out
Mike Frysinger [Sat, 16 Dec 2006 19:18:08 +0000 (19:18 -0000)]
actually scrub it, dont just comment it out

17 years agostart pulling in top level Rules.mak so we get access to import CFLAGS/LDFLAGS/CPPFLAGS
Mike Frysinger [Sat, 16 Dec 2006 19:13:21 +0000 (19:13 -0000)]
start pulling in top level Rules.mak so we get access to import CFLAGS/LDFLAGS/CPPFLAGS

17 years agoscrub prototype so we dont conflict with the pthread header
Mike Frysinger [Sat, 16 Dec 2006 19:12:45 +0000 (19:12 -0000)]
scrub prototype so we dont conflict with the pthread header