It's not necessary to remove the set -u. The false was crashing the compile.
Search Criteria
Package Details: gcc45 4.5.4-3
| Git Clone URL: | https://aur.archlinux.org/gcc45.git (read-only) |
|---|---|
| Package Base: | gcc45 |
| Description: | The GNU Compiler Collection (4.5.x) |
| Upstream URL: | http://gcc.gnu.org |
| Licenses: | |
| Conflicts: | |
| Submitter: | severach |
| Maintainer: | severach |
| Last Packager: | severach |
| Votes: | 0 |
| Popularity: | 0.000000 |
| First Submitted: | 2016-01-05 00:33 |
| Last Updated: | 2016-08-06 06:26 |
Dependencies (12)
- binutils (binutils-git, binutils-hjl-git, binutils-tune-bfd-hash)
- cloog (cloog-git)
- elfutils
- glibc
- gmp
- isl (isl-git, isl15)
- libmpc
- mpfr
- ppl
- bison (byacc-bison) (make)
- flex (flex-android) (make)
- setconf (make)
Required by (1)
- sdlmame-0137 (make)
Sources (3)
- gcc-hash-style-both.patch
- gcc_pure64.patch
- http://www.mirrorservice.org/sites/sourceware.org/pub/gcc/releases/gcc-4.5.4/gcc-4.5.4.tar.bz2
Latest Comments
severach commented on 2016-06-07 16:00
biinari commented on 2016-06-07 12:53
`set -u` does not seem to play nice with makepkg, resulting in this error masking other errors in prepare/build/package:
/usr/bin/makepkg: line 729: logpipe: unbound variable
Also, on the case line for gcc_pure64.patch, there is a `false` statement causing the build to fail.
And, as per gcc49, we need to give `--build=${CHOST}` to configure.
I have created a patch to the PKGBUILD at https://gist.github.com/biinari/8f731cec631464ef8a045205e87a347e