Commit Graph

276 Commits

Author SHA1 Message Date
mrg
d35849d0e1 merge GCC 5.3. round 1. this probably won't work :-)
libsanitizer likely needs to be re-ported, though i've done most of
the updating we'll need i think already.
2016-01-24 09:43:30 +00:00
mrg
4d5abbe83f import GCC 5.3.0. see these urls for details which are too large to
include here:

	http://gcc.gnu.org/gcc-4.9/changes.html
	http://gcc.gnu.org/gcc-5/changes.html

(note that GCC 5.x is a release stream like GCC 4.9.x, 4.8.x, etc.)


the main issues we will have are:

The default mode for C is now -std=gnu11 instead of -std=gnu89.

ARM:
The deprecated option -mwords-little-endian has been removed.
The options -mapcs, -mapcs-frame, -mtpcs-frame and -mtpcs-leaf-frame
 which are only applicable to the old ABI have been deprecated.

MIPS:
The o32 ABI has been modified and extended. The o32 64-bit
 floating-point register support is now obsolete and has been removed.
 It has been replaced by three ABI extensions FPXX, FP64A, and FP64.
 The meaning of the -mfp64 command-line option has changed. It is now
 used to enable the FP64A and FP64 ABI extensions.
2016-01-24 06:05:37 +00:00
mrg
93abc59558 protect the entire contents of regsub.c with extern "C".
this fixes the tools build, at least.

also, remove the no longer needed cdefs.h from one of these copies.
2016-01-21 04:55:40 +00:00
christos
c424ac81dd make regasub a c decl. 2016-01-20 17:08:27 +00:00
christos
168ddffc8a sync with gcc.old (catch up with __RCSID removal and libc sync) 2016-01-20 15:13:33 +00:00
christos
10dfc2597e ifdef __RCSID 2016-01-16 19:28:36 +00:00
wiz
40fc864a9c Add missing word in comment. 2016-01-11 22:41:48 +00:00
christos
3965f5d7b6 PR/50646: Use the c99 cabs math builtins directly, because cabs and cabsf
have broken ABI's can cabsl does not exist. The correct(?) probably fix is
to change the cabs builtins to point to the c99 variants directly...
XXX: pullup-7 and the same file from the gcc.old tree.
2016-01-11 16:52:53 +00:00
christos
ee2af19067 add -fdebug-regex-map=regex=subst which works like sed -e s/regex/subst/
to aid with /usr/obj remapping for MKREPRO
2016-01-09 02:00:14 +00:00
christos
37ae74be41 For reproducible builds, allow the source of the map to be set from the
environment, so that DW_at_producer is the same no matter what the actual
underlying source and destination directories are. See:
    https://gcc.gnu.org/ml/gcc-patches/2015-12/msg01168.html
2015-12-22 16:48:44 +00:00
christos
4fe0054f39 Fix logic bug where __cpuid gets called when __get_cpuid_max() returns 0.
This happens when __level == 0, from libgcc/config/i386/cpuinfo.c#285.
This breaks early i486 CPUs which don't have cpuid
XXX: pullup-7
2015-11-22 20:47:05 +00:00
christos
b09d094867 We've never supported the executable stack note, and none of the other
archs have it.
2015-11-08 17:11:55 +00:00
christos
bb5915f98c Fix inconsistencies with GNU-stack note:
- always use __ELF__ && __linux, not just __ELF__ or just __linux__
- remember to pop back to the previous section where it is missing
XXX: need to file this bug with the GNU folks.
2015-11-07 16:53:07 +00:00
gdt
fc89a2852b Use -f with cp.
When the source tree is 444 (as should be unremarkable), cp results in
object files that are 444, which when cp'd again without -f result in
an error.
2015-11-02 00:51:18 +00:00
christos
d8e986fab6 Introduce a syslog format that accepts %m. Stop accepting %m in printf
like formats. Support for this is detected via the __syslog_attribute__
macro.
2015-10-14 16:05:14 +00:00
jakllsch
23d6404318 Correct temporary directory preference order in libiberty's choose_tmpdir().
Because it is intended to be persistent, /var/tmp is about the worst possible
choice for temporary files for most users of libiberty.  /tmp works better,
because the the defined semantics of /tmp allow for a non-persistent tmpfs
to be used.  This should improve performance when /tmp is a tmpfs and it is
difficult or impossible to have an environment variable or command line -pipe
flag passed to every piece of the toolchain.
2015-08-10 15:45:40 +00:00
mrg
3008502ee0 merge GCC 4.8.5 (final round) 2015-06-24 14:51:17 +00:00
mrg
11d0566779 merge GCC 4.8.5 (round two) 2015-06-24 14:29:51 +00:00
mrg
788f3c6afe merge GCC 4.8.5 cornflakes (round one). 2015-06-24 14:28:27 +00:00
mrg
7e9e220a26 initial import of GCC 4.8.5 sources. these GCC PRs (at least) are
fixed in this release:

64882 46102 58123 65409 59626 61058 65680 64487 65721 65727
52306 64768 56273 57653 59990 60656 61634 63844 64199 64493
64495 65549 57059 57569 57748 58369 62642 63608 64037 65550
65693 65220 45187 64409 61977 64513 64634 65368 66233 66470
65072 65327 65543 65279 63593 65063 65518 66123 66140 59016
64479 64557 64979 66215 66275 66481 39423 64766 63733 52714
60898 61138 63744 57023 39722 45402 52579 52664 60718 62044
59513 35330 37440 43701

see:
  https://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.8.5
for more details.
2015-06-24 14:02:39 +00:00
matt
b766b2b592 Change CC1_SPEC andd CC1PLUS_SPEC to supply -fPIC by default if none of
m32 mrelocatable fpie fPIE fpic fPIC fno-pie fno-pic are supplied.
2015-06-24 01:29:33 +00:00
matt
7593e659ff By changing rs6000_elf_reloc_rw_mask to return 3 if using a powerpc64 ABI
on NetBSD, we basicly eliminate the only difference between have -fPIC or
not.  This enables us to turn on MKPICLIB for powerpc64.
2015-06-23 01:37:13 +00:00
matt
6b8427896d Add the historical __OCTEON__ builtin_define along with __mips_popcount which
is defined if the platform has the dpop/pop instructions.
2015-05-29 19:37:27 +00:00
matt
e3b08ce148 Make MULTILIBS work for oabi 2015-05-27 15:33:41 +00:00
skrll
640e856ca3 Fix up TARGET_ENDIAN_DEFAULT for BE oabi 2015-05-25 21:18:29 +00:00
prlw1
9521ae7bfd Fix typos 2015-04-28 09:48:30 +00:00
joerg
66e8846b8d Redo __aeabi_atexit by shuffling the arguments and dropping the code in
libstdc++.
2015-04-19 18:15:26 +00:00
martin
b40e9b5326 Add evil hack to help mknative untill we finish userland 2015-04-17 12:29:49 +00:00
martin
4ae85d6ddd Add missing cast 2015-04-17 12:29:00 +00:00
bouyer
8bc54b22df Use the same TRY_EMPTY_VM_SPACE as mips and vax for arm, which also
has VM_MAXUSER_ADDRESS set to 2GB. Fixes "fatal error: had to relocate PCH"
from g++ building wxGTK28 or wxGTK30.
For details see https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58379

thanks to Martin Husemann for the hint.
2015-04-03 17:25:42 +00:00
dennis
f334ce9265 Make the small [U]INT_LEAST?_TYPE definitions match the 32 bit compiler 2015-04-02 16:49:09 +00:00
matt
6d5bd4db35 Update to latest gcc changes. Default NetBSD to the medany cmodel. 2015-04-01 06:35:01 +00:00
matt
019155557a Add RISCV bits 2015-03-31 17:58:30 +00:00
matt
9e7510ab24 Update RISCV tls assembly tests 2015-03-31 17:57:40 +00:00
matt
163496bc0e Deal with NetBSD using dwarf EH 2015-03-31 17:18:35 +00:00
matt
11b22aab18 Make gcc emit the CPP macros for integer types. 2015-03-29 00:42:51 +00:00
matt
9b59c19e65 Make sure the RISCV compiler defaults to FDIV enabled. 2015-03-28 00:47:17 +00:00
matt
1523fd3ab8 Update gcc to use new ABI for RISCV 2015-03-27 01:51:34 +00:00
christos
4fc6ba4368 fix for PIC code. 2015-01-25 20:06:46 +00:00
matt
2eb84c7288 Make sure POWERPC_NETBSD is defined. 2015-01-22 08:35:05 +00:00
matt
9a22733cae Don't enable string instructions on NetBSD if -Os is used. 2015-01-20 07:12:41 +00:00
skrll
b4b886e5d8 Fix gcc PR target/64479 [SH] wrong optimization delayed-branch 2015-01-15 08:42:14 +00:00
mrg
94b9bff57a merge GCC 4.8.4. 2015-01-10 01:06:41 +00:00
mrg
6012017c82 import GCC 4.8.4 release. this fixes at least these GCC PRs:
fortran/61407
	sanitizer/64265
	tree-optimization/64269
	middle-end/64225
	tree-optimization/61686
	bootstrap/64213
	rtl-optimization/64037
	target/50751
	rtl-optimization/64037
	target/59593
	target/59593
	c++/56493
	target/64115
	middle-end/64111
	middle-end/64067
	rtl-optimization/63659
	libgomp/61200
	tree-optimization/61969
	tree-optimization/62031
	tree-optimization/63379
	tree-optimization/63605
	middle-end/63665
	target/60111
	target/63673
	target/63947
	tree-optimization/62167
	tree-optimization/63841
	ipa/63838
	c++/63455
	c++/63415
	c++/56710
	c++/58624
	preprocessor/60436
	target/55351
	fortran/63938
	libgomp/61200
	libstdc++/63840
	libstdc++/61947
	libstdc++/59603
	target/56846
	libstdc++/57440
2015-01-10 00:47:09 +00:00
dennis
298eb8b514 Apply a fix, extracted from the UCB RISC-V gcc sources, for the (known)
problems with compiling atomics for the riscv64 target with gcc.  Remove
the now-unneeded workarounds for the 2 files in libasan that were formerly
failing to compile with the optimizer turned on.
2014-12-22 21:15:48 +00:00
christos
6eb7640629 make alpha compile 2014-10-23 19:03:38 +00:00
christos
3ce54f1012 use the appropriate unwind.h header 2014-10-23 17:50:24 +00:00
christos
a62fb610ef stop playing type games, and use the proper type for the placement new operator 2014-10-23 17:50:08 +00:00
christos
9512ff2aab simplify and add more architectures. 2014-10-23 00:25:54 +00:00
christos
9b7e305c80 set the proper asan variable 2014-10-22 23:17:24 +00:00