similar to: [PATCH] support the gcc parameter --param in klcc

Later versions of gcc include an option in the form of --param=name=value. This patch adds support for this and continues to support the --param name=value option. Signed-off-by: Mike Pagano <mpagano at gentoo.org> --- klcc/klcc.in | 15 +++++++++++---- 1 files changed, 11 insertions(+), 4 deletions(-) diff --git a/klcc/klcc.in b/klcc/klcc.in index 711a832..5d00bfd 100644 ---

Here's a patch to klcc.in from klibc-0.202. The context: I have an application, managed with GNU automake/autoconf. Packagers should be able to build the application with klibc if available, with plain gcc otherwise, as follows: $ cd ~/klibc-0.202 $ ... make linux symlink $ make bindir=$HOME/local/bin \ mandir=$HOME/local/man \ INSTALLDIR=$HOME/local \ SHLIBDIR=$HOME/local/shlib \

Signed-off-by: Thorsten Glaser <tg at mirbsd.org> --- klcc/klcc.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/klcc/klcc.in b/klcc/klcc.in index 43d0984..e03bf3c 100644 --- a/klcc/klcc.in +++ b/klcc/klcc.in @@ -136,7 +136,7 @@ while ( defined($a = shift(@ARGV)) ) { } else { die "$0: unknown option: $a\n"; } - } elsif ( $a =~

Commit-ID: 7752eb456addbd1973e9c6b609294ccef89a65a2 Gitweb: http://git.kernel.org/?p=libs/klibc/klibc.git;a=commit;h=7752eb456addbd1973e9c6b609294ccef89a65a2 Author: Ben Hutchings <ben at decadent.org.uk> AuthorDate: Thu, 20 Aug 2020 21:48:07 +0100 Committer: Ben Hutchings <ben at decadent.org.uk> CommitDate: Thu, 20 Aug 2020 21:50:09 +0100 [klibc] klcc: Treat CC, LD, STRIP

sscanf can take an integer which specifies the maximum field width. In this test, the field width is 1 and the input is a negative number so the test fails. This patch increases the maximum field width to 2. Signed-off-by: Mike Pagano <mpagano at gentoo.org> --- usr/klibc/tests/sscanf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/usr/klibc/tests/sscanf.c

Currently, in the entire klibc bash is only used to identify the path of the perl binary. It is doing so using the bash built-in 'type' function, which is POSIX compliant according to [0], but the option -c is not. By using `command -v` instead, we achieve the same result, in a POSIX compliant manor [1], potentially removing the unneeded bash dependency. 0

There's a problem in klibc 1.0.12: klcc -c t.c && klcc t.o && ./a.out works, but klcc t.c && ./a.out fails because open3() wants to close its input, and you can do that only once to STDIN. Attached patch makes "klcc t.c" work again. This might break the klcc from stdin option, but I can't tell because I've got no idea how it's supposed to

Stop makeklcc.pl from hardcoding paths of cc, ld and strip in the klcc script it generates. Using hardcoded paths is generally a bad idea. First, the whole idea of $PATH is that you don't need to hardcode paths. Second, klcc is a Perl script but my hardcoding the paths you make it less portable. Signed-off-by: Martin Michlmayr <tbm@cyrius.com> --- klibc-1.1.16/klcc/makeklcc.pl~

Halo, I was wondering to no avail how to to use the klcc wrapper in order to build udev and module-init-tools. Does klibc has an install target when is embedded in the kernel tree? Is klcc available? .bill

Package: klibc Version: 1.5.20-1 Severity: normal Tags: patch User: ubuntu-devel at lists.ubuntu.com Usertags: origin-ubuntu natty ubuntu-patch Hi maks, Since the Ubuntu linux-libc-dev is now installing its asm headers to the multiarch /usr/include/<triplet>/ directory to allow co-installability, I've pushed the attached patch to Ubuntu so klcc knows to include this path when searching

Sorry for the previous HTML message. This time in plain text. sscanf can take an integer which specifies the maximum field width. In this test, the field width is 1 and the input is a negative number sothe test fails. This patch increases the maximum field width to 2. Signed-off-by: Mike Pagano <mpagano at gentoo.org> --- usr/klibc/tests/sscanf.c | 2 +- 1 file changed, 1 insertion(+), 1

Commit-ID: 686f1931b7dc102c32b59db638a3082d5c2251e5 Gitweb: http://git.kernel.org/?p=libs/klibc/klibc.git;a=commit;h=686f1931b7dc102c32b59db638a3082d5c2251e5 Author: maximilian attems <max at stro.at> AuthorDate: Mon, 10 Jun 2013 18:15:43 -0400 Committer: maximilian attems <max at stro.at> CommitDate: Wed, 21 Aug 2013 11:50:43 +0200 tests: Fix sscanf integer tests Drop the

Commit-ID: 95a3123c94abc28c65f33c9589693dde9f140fe1 Gitweb: http://git.kernel.org/?p=libs/klibc/klibc.git;a=commit;h=95a3123c94abc28c65f33c9589693dde9f140fe1 Author: Ben Hutchings <ben at decadent.org.uk> AuthorDate: Sat, 29 Aug 2020 22:29:52 +0100 Committer: Ben Hutchings <ben at decadent.org.uk> CommitDate: Sat, 29 Aug 2020 22:53:10 +0100 [klibc] Kbuild, klcc: Support

Commit-ID: 23273aa05bdc564eb8ff0565d9df9c3edfffaee9 Gitweb: http://git.kernel.org/?p=libs/klibc/klibc.git;a=commit;h=23273aa05bdc564eb8ff0565d9df9c3edfffaee9 Author: Olliver Schinagl <oliver at schinagl.nl> AuthorDate: Sat, 25 Jul 2020 10:59:01 +0200 Committer: Ben Hutchings <ben at decadent.org.uk> CommitDate: Sun, 26 Jul 2020 00:16:13 +0100 klcc: Remove the need for bash

On Sat, 2020-07-25 at 10:59 +0200, Olliver Schinagl wrote: > Currently, in the entire klibc bash is only used to identify the path of > the perl binary. It is doing so using the bash built-in 'type' function, > which is POSIX compliant according to [0], but the option -c is not. > > By using `command -v` instead, we achieve the same result, in a POSIX > compliant manor

Commit-ID: f24c4d725f5b178d1027dca14203f57e8e78a34d Gitweb: http://git.kernel.org/?p=libs/klibc/klibc.git;a=commit;h=f24c4d725f5b178d1027dca14203f57e8e78a34d Author: Ben Hutchings <ben at decadent.org.uk> AuthorDate: Tue, 19 Feb 2019 02:31:39 +0000 Committer: Ben Hutchings <ben at decadent.org.uk> CommitDate: Tue, 19 Feb 2019 02:31:39 +0000 [klibc] klcc: Enable stripping

Ben Hutchings <ben at decadent.org.uk> wrote: > On Sat, 2020-07-25 at 10:59 +0200, Olliver Schinagl wrote: >> Currently, in the entire klibc bash is only used to identify the path of >> the perl binary. It is doing so using the bash built-in 'type' function, >> which is POSIX compliant according to [0], but the option -c is not. >> >> By using

Trying to use klcc to compile an application I was getting gcc: no input files. Tracked it down to filename2lang having and extra space, causing .S files to be labeled "obj". Signed-off-by: Milton Miller <miltonm at bga.com> --- There is probably another bug that we called gcc without any files when we were called with what we parsed as an obj file. ---

Some packages want to pass input to gcc via STDIN (iptables for example) This patch lets klcc accept '-' as a valid input filename, prevents it from setting -x in that case and changes the invocation of gcc from system() to open3() so that gcc's STDIN can be connected to klcc's STDIN. Daniel Thaler -------------- next part -------------- --- klibc-1.0.8/klcc.in.old 2005-05-20

Hi Currently I am trying to get device-mapper and dmraid ported to compile against klibc. So far there was no real issues, as the only thing they used was fscanf and feof not available presently in klibc (but easy to work around). Other than that there are really no way to easily check for installed klibc's other than trying to mangle the output of klcc (from adjusted configure.in): -----