Age | Commit message (Collapse) | Author |
|
debug information, gdb will either see calltrap or alltraps in the
stack trace of a kernel core file. To make the gdb backtrace command
work without debug information in the amd64 kernel, add the same
special case for alltraps as for calltrap.
OK miod@
|
|
OK deraadt@ miod@
|
|
on slower platforms. Idea from espie@, OK espie@ deraadt@ guenther@
|
|
as an ersatz for -W. Now that more and more third-party software assumes the
compiler supports -Wextra, this is definitely worth doing.
|
|
gcc 3 no longer defines this for us.
|
|
members to 64bit types. Assign new syscall numbers for (almost
all) the syscalls that involve the affected types, including anything
with time_t, timeval, itimerval, timespec, rusage, dirent, stat,
or kevent arguments. Add a d_off member to struct dirent and replace
getdirentries() with getdents(), thus immensely simplifying and
accelerating telldir/seekdir. Build perl with -DBIG_TIME.
Bump the major on every single base library: the compat bits included
here are only good enough to make the transition; the T32 compat
option will be burned as soon as we've reached the new world are
are happy with the snapshots for all architectures.
DANGER: ABI incompatibility. Updating to this kernel requires extra
work or you won't be able to login: install a snapshot instead.
Much assistance in fixing userland issues from deraadt@ and tedu@
and build assistance from todd@ and otto@
|
|
- provide 88110 syntactic sugar for the control register names
- correctly handle -m options
- and a minor fix to allow the register prefix to correctly be recognized in
front of the condition codes (only needed for SVR4)
|
|
httpd correction and ok kettenis@
|
|
|
|
matches the behaviour of gcc >= 4.3 and clang
ok miod@
|
|
considerations.
|
|
Change the logic depending upon COMPILER_VERSION everywhere, to assume gcc4
is the norm and to explicitely test for gcc3 when a different behaviour
is required.
No functional change intended. Be sure to `make install' in share/mk before
attempting to do anything.
|
|
|
|
with an optimization pass which attemps to identify and remove dead stores.
Unfortunately, this logic assumes that a given rtl MEM reference can only
refer to register contents and values, but not further dereference memory.
This is not true of all addressing modes on vax, and this causes gcc 3 to
consider as dead stores everything which can be expressed as *offset(register)
in vax assembly. The canonical example of this is linked list insertion or
removal, with instructions such as "item->prev->next = item->next" being
wrongly optimized out, and cause double frees or other funny linking problems
later on.
Add a knob to disable the troublesome part of that optimization pass, and
only enable this knob on vax.
This gives a native vax gcc 3 compiler able to rebuild the world.
|
|
considerations.
|
|
considerations.
|
|
ok kettenis@
|
|
|
|
value can not be represented by a double (unless compiling with -mg); instead,
use a lower limit value. This does not impact the behaviour since both sides
of the test are supposed to produce the same result, albeit in different ways.
ok millert@ espie@
|
|
|
|
|
|
|
|
NetBSD.
|
|
|
|
From Joerg Wunsch in GCC PR 23479, under the GPLv2.
This is required to build the i965 backend with newer versions of mesa.
ok kettenis@ espie@ miod@
|
|
considerations.
|
|
directory if it doesn't exist.
ok sthen@, espie@
|
|
block where it had been mistakenly put in the first place (in r1.28).
|
|
|
|
|
|
|
|
instruction, before testing whether it can be built with `or.u'; this
allows further optimization.
condition_value(): handle ORDERED and UNORDERED condition codes.
print_operand(): remove support for no longer used `w' qualifier.
|
|
inline assembly, tb/tbnd instructions.
Don't use `r' qualifier for "register_operand", it's redundant.
|
|
because that resulted in corrupt output.
This is a verbatim backport of the following bugfix from upstream git master:
https://rt.cpan.org/Public/Bug/Display.html?id=77465
https://github.com/mrallen1/Pod-Perldoc/commit/6e1541d0bcb74a7b2b9ee3235d57953fb800bb67
Do not take the comment in the source code too literally.
It doesn't really explain the problem well.
OK sthen@ espie@ millert@
|
|
|
|
Will send this upstream as well.
OK millert@ bluhm@ stsp@
|
|
generator)
writing a commitid is currently disabled
ok deraadt
|
|
|
|
|
|
|
|
FreeBSD).
|
|
This makes gcc -fprofile-arcs work again.
Regression reported by Azwaw OUSADOU on tech@, thanks for noticing and sorry
for the delay fixing this.
|
|
same solution from FreeBSD.
Don't even try to build it at the moment. This is commited so people can
work on it in tree.
|
|
Fix found upstream.
ok miod@, martynas@
|
|
make -j might rebuild those spuriously when they end up being out of date...
okay millert@
|
|
dynamically linked binaries. Borrowed from m68k with trivial changes.
ok kettenis@
|
|
Main features:
- md constraints rewritten in RTL
- md predicaties rewritten in RTL
- md va_arg switched to gimple
- abort() calls replaced with gcc_assert() or gcc_unreachable() for better
diagnostics
- support for non-ELF systems completely removed
Missing:
- conversion of the pipeline information from define_function_unit to
define_automata not done yet (thus pipeline information currently removed)
Known regressions against 3.3.6 so far:
- no stack protector support yet
- __builtin_setjmp doesn't restore the frame pointer correctly upon return
from __builtin_longjmp
- at least one case of optimization error when delay slots are not disabled.
- libgcc is only built -fPIC, instead of static/fpic/fPIC.
|
|
incomplete equivalent. The ugly BLK<->{QI,HI,SI,DI} load and store insns in
m88k.md can thus get removed, and there was much rejoicing (as well as fewer
warnings when building).
|
|
|
|
issue it fixes will be fixed anyway once the move to binutils-2.17 happens.
xenocara breakage discovered by matthieu@
|