TODO list for the uClibc 1.0.0 release:
-------------------------------------------------
+ *) glob / fnmatch tests fail
*) Finish hiding uClibc internal symbols from our exported namespace.
*) Explicity add some sortof CONFIG_LINUX_2_2_ABI, CONFIG_LINUX_2_4_ABI
and CONFIG_LINUX_2_6_ABI type options, rather than having the abi
Arch specific TODO:
-------------------------------------------------
alpha:
- signal handling is all screwed; common linux versions do not work for us,
- need to import all the stuff from glibc ...
- _init/_fini hangs signal-ed (assert test)
- output from cloned children seems funky (unistd/clone test)
+
ldso is not implemented
arm:
update crt1 code again for nommu (http://bugs.uclibc.org/view.php?id=538)
--- /dev/null
+/* Copyright (C) 1998, 2003, 2004 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Richard Henderson <rth@cygnus.com>, 1998
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#include <features.h>
+#include <sys/syscall.h>
+#include <asm/regdef.h>
+#include <asm/pal.h>
+
+/* On Alpha we desparately want to avoid having to issue an imb. Ordinarily
+ the kernel would have to issue one after setting up the signal return
+ stack, but the Linux rt_sigaction syscall is prepared to accept a pointer
+ to the sigreturn syscall, instead of inlining it on the stack.
+
+ This just about halves signal delivery time. */
+
+ .text
+
+.globl __syscall_rt_sigaction
+.align 4
+.ent __syscall_rt_sigaction, 0
+__syscall_rt_sigaction:
+ .frame sp,0,ra,0
+ ldgp gp,0(pv)
+ .prologue 1
+
+ beq a1, 0f
+ ldl t0, 8(a1) # sa_flags
+
+ /* The unwinder will subtract one from the return address when
+ attempting to find the call instruction that led us here.
+ Since we didn't get here via a normal call, if we do nothing
+ we would pick up the wrong symbol and the wrong FDE. Account
+ for this by adding a nop to the start of the function and
+ then skipping it here by adding 4. */
+ ldah a4, __syscall_sigreturn+4(gp) !gprelhigh
+ ldah t1, __syscall_rt_sigreturn+4(gp) !gprelhigh
+ lda a4, __syscall_sigreturn+4(a4) !gprellow
+ lda t1, __syscall_rt_sigreturn+4(t1) !gprellow
+ and t0, 0x40, t0 # SA_SIGINFO
+ cmovne t0, t1, a4
+
+0: ldi v0, __NR_rt_sigaction
+ callsys
+ bne a3, $error
+ ret
+
+$error:
+ jmp zero,__syscall_error
+
+.end __syscall_rt_sigaction
+
+__syscall_sigreturn:
+ nop
+ mov sp, a0
+ ldi v0, __NR_sigreturn
+ callsys
+ .size __syscall_sigreturn, .-__syscall_sigreturn
+ .type __syscall_sigreturn, @function
+
+__syscall_rt_sigreturn:
+ nop
+ mov sp,a0
+ ldi v0,__NR_rt_sigreturn
+ callsys
+ .size __syscall_rt_sigreturn, .-__syscall_rt_sigreturn
+ .type __syscall_rt_sigreturn, @function