 Ralf S. Engelschall
|
5a2ea630df
modifying package: gcc-3.3.1 20030808 -> 20030809
|
22 years ago |
|
Ralf S. Engelschall
|
9bdcb61169
upgrading package: gcc 3.3 -> 3.3.1
|
22 years ago |
 Michael Schloh von Bennewitz
|
e3736b8ff6
RT#220, correct group names
|
22 years ago |
|
Ralf S. Engelschall
|
57dcbf7729
achieve robust subshell handling by consistently using '(...) || exit 0' constructs to make sure the scripts fail if a subshell fails
|
22 years ago |
|
Ralf S. Engelschall
|
acda31d548
modifying package: gcc-3.3 20030721 -> 20030722
|
22 years ago |
|
Ralf S. Engelschall
|
54eb7449b6
speclint police: achieve semantical consistency for patch references
|
22 years ago |
 Thomas Lotterer
|
8cb96928b0
modifying package: gcc-3.3 20030702 -> 20030717
|
22 years ago |
|
Ralf S. Engelschall
|
67f9c78a59
cleanup copyright part by using just the main company name of Cable & Wireless and adding the URL to the global website
|
22 years ago |
|
Ralf S. Engelschall
|
6e71b425a1
pay attention to the multilib support in gcc and move also libs in subdir(s). This is mainly important under Solaris where 32 and 64 bit versions of libs are build and installed
|
22 years ago |
|
Ralf S. Engelschall
|
795b2642b6
make sure bounds and spp are not used at the same time and make sure the build terminated if the spp patching failed
|
22 years ago |
|
Ralf S. Engelschall
|
e6eae14d95
add optional support for IBM Stack-Smashing-Protection patch
|
22 years ago |
|
Ralf S. Engelschall
|
55f917a1e8
Unfortunately, the GCC documentation and the implementation diverge dramatically: BOOT_CFLAGS is not for stage1 (booting), but for stage2, stage3, etc. And LIBCFLAGS are used for libiberty, but no BOOT_LIBCFLAGS exists (BOOT_LIBCXXFLAGS are not required, btw). To allow GCC to build correctly with a non-GCC vendor compiler, we no longer set LIBCFLAGS at all (because libiberity build would fail) and we use CFLAGS for the stage1 and BOOT_CFLAGS for stage2, stage3, etc. So, beastie GCC, DWIM now, please!
|
22 years ago |
|
Ralf S. Engelschall
|
8e1db1d0df
finally fix the build flags handling
|
22 years ago |
|
Thomas Lotterer
|
a1d0b92ef5
modifying package: gcc-3.3 20030614 -> 20030625
|
22 years ago |
|
Michael van Elst
|
59ac152580
fix building on non-gcc hosts
|
22 years ago |
|
Ralf S. Engelschall
|
16e6ba3e5f
upgrading package: gcc 3.2.3 -> 3.3
|
22 years ago |
|
Ralf S. Engelschall
|
7161673984
modifying package: gcc-3.2.3 20030427 -> 20030430
|
22 years ago |
|
Ralf S. Engelschall
|
33bc9e4c27
add optional bounds checking patch
|
22 years ago |
|
Ralf S. Engelschall
|
952816f343
upgrading package: gcc 3.2.2 -> 3.2.3
|
22 years ago |
|
Ralf S. Engelschall
|
198e060687
try to remove hopefully obsolete (and already making problems) patch for disabling the usage of GCC includes
|
22 years ago |
|
Ralf S. Engelschall
|
c33f4a18b0
be more precise in matching
|
23 years ago |
|
Ralf S. Engelschall
|
f33ede2650
make flag determination more flexible and especially add more optional optimization
|
23 years ago |
|
Michael van Elst
|
e3325ff0b4
patch for Solaris/x86
|
23 years ago |
|
Ralf S. Engelschall
|
b0d2348b03
Completely work off our GCC packaging. This especially now uses the (now finally working) DESTDIR support in GCC and cleans up the parameter determinations. As a side-effect the packaging shrunk by about 1KB, too.
|
23 years ago |
|
Ralf S. Engelschall
|
bdedcee0d9
upgrading package: gcc 3.2.1 -> 3.2.2
|
23 years ago |
|
Thomas Lotterer
|
aa0fcf0b0f
try to fix the patch according to FreeBSD gcc32 port in order to solve varargs problems
|
23 years ago |
|
Ralf S. Engelschall
|
c258027d7b
Include official patch for FreeBSD 5 from gcc.gnu.org CVS repository: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ginclude/stddef.h.diff?r1=1.15&r2=1.16 This fixes building under latest FreeBSD 5.0-CURRENT.
|
23 years ago |
|
Ralf S. Engelschall
|
7a694d853e
upgrade to new %option world order
|
23 years ago |
|
Ralf S. Engelschall
|
4a73168b54
bump of Copyright messages for forthcoming years 2003
|
23 years ago |
|
Ralf S. Engelschall
|
ab464b683d
switch to %option(s) and remove dynamic default for binutils because this makes trouble and is inconsistent anyway
|
23 years ago |
