| Age | Commit message (Collapse) | Author |
|---|
|
ok miod
|
|
building cvs fails after a forgotten 'make obj', but the problem is tough
to spot because the files which cause a problem are hidden from 'cvs up'
output (even with 'cvs up -I !') because they're listed in .cvsignore.
|
|
for ages by now.
|
|
|
|
ok miod@, jsg@
|
|
|
|
|
|
|
|
for unwind-dw2.c fails to build otherwise.
|
|
|
|
unable to find a register to spill in class 'R0_REGS'
ICE when compiling PIC code. Similar to the MI targhooks.c change.
|
|
ok miod@ espie@
|
|
|
|
|
|
|
|
|
|
written by guenther@ about a year ago.
discussed with him and millert@
|
|
helpful. Just provide the TID instead.
|
|
|
|
with _REENTRANT. Also remove undef _POSIX_THREADS in phread.h.
From brad@comstyle.com. okay guenther@
|
|
instead of global with gcc4 and this ends up breaking things in hard
to debug ways. So move the definition to csu instead of libgcc.
ok miod@
|
|
so we can use the existing sh libkern functions.
ok miod@
|
|
|
|
|
|
|
|
|
|
|
|
version of GDB compared to what's currently upstream.
|
|
Only amd64 for now; more architectures will follow.
|
|
totally fucked up, effectively leading to inverted behaviour of the -fident and
-fno-ident flags (while keeping the .ident string in by default). Revert it
back to normal and default to -fno-ident.
ok millert@
|
|
|
|
ecoff moot blocks on sgi.
|
|
a lifesaver for R4000 and R4400 operation; without these, every int->long
promotion occuring shortly after an integer multiplication (such as an array
element access when the array item size is not a power of two) loses horribly,
and panic^Whilarity ensues.
This mostly causes assembly insns to be shuffled, but almost no code size
growth and no noticeable performance hit on processors which do not need
these insn placement workarounds.
|
|
- don't ignore "core", it gets in the way when we import certain things
- ignore "*.core" to match the actual filenames we use for core dumps
suggested by landry@
- ignore ".git" to avoid accidental imports of the directory when used
locally "on top of" cvs
various versions ok'd by landry@ todd@ espie@ phessler@ dcoppa@
|
|
Found by Brad.
Also, consider stpcpy an unsafe builtin and prevent it from being optimised
away.
ok miod@
|
|
occurrences to get_line().
Based on a diff from Jan Klemkow <j-dot-klemkow-at-wemelug-dot-de> to tech.
|
|
kprintf supports %z, doesn't support old ddb %n/%r%z anymore, and check of
%b arguments have to occur on the next argument, like, duh.
These changes will eventually allow for kernel to be compiled without
-Wno-format, but some more casts or type changes are necessary, first.
Discussed about six months ago at s2k11, time to put this in, so that I have no excuse not to work on the format fixes diffs.
|
|
use of awk. Slightly faster and works around an apparent namei or
buffer cache related bug on arm. Requested and OK deraadt@
|
|
|
|
okay fgs@, millert@
|
|
ok miod@
|
|
|
|
|
|
C++98, so clang++ rightfully complains about it when in C++98 mode (the
default).
Found while investigating the other clang issue, with helpful input from
espie@.
ok jsg@
|
|
without barfing about conflicting declarations.
Issue reported by Amit Kulkarni, investigated by matthew@ and kettenis@.
ok kettenis@
|
|
memory usage optimization, because this code is just bogus and causes
access-beyond-allocated-size SIGSEGV in the wild.
This is part of a larger diff (being worked on) backporting the removal of
``permanent_obstack'' completely from gcc, as done by the gcc folks in 3.x
after they came to their senses (or got tired of dissecting coredumps on
platforms with strong malloc implementations).
Only affects m68k, m88k and vax; collecting dust in the m88k tree, commit
triggered by sebastia@ via espie@ as this fixes the devel/gperf port on vax.
|
|
and change ASM_OUTPUT_ALIGNED_COMMON to output an explicit .align directive,
as gas will not compute it from the size (despite what comments elsewhere in
gcc code say).
This effectively makes __attribute__ ((aligned (N))) constructs work as
intended for .bss objects.
ok kettenis@
|
|
word boundaries, and enable the builtins.c `one-byte memcpy' workaround.
|
|
size 1 (the size being known at compile-time) into an inline mempcpy()
expansion, which will in turn expand into a byte load and store operation.
This expansion loses precious address alignment information at some point
(because everybody knows that you can read a byte from any address, right?),
and this loses bigtime on strict alignment platforms which lack the ability
to accesse bytes directly, such as alpha (unless compiling with -mbwx and
runnning on a BWX-capable cpu).
An example of such call with size 1 is lib/libkvm/kvm.c line 780.
|
|
different libstdc++.so majors. ok deraadt@
|
