- CRUX-ARM - lists.crux-arm.nu

RC6 results and RC7 plans
by Jose V Beneyto 02 Sep '10

02 Sep '10

Hi! I tested the noeabi flavour of RC6 in my HP Jornada 720 and qemu, and both of them worked fine! I think core is now more stable and the best ever, but as we talked many times at irc channel we should add a package manager like pkg-get, so after the installation people could continue completing their systems with hosted packages instead of compiling new ones. This would be helpful for slow devices, or devices with less than 64MB of ram, where compile a port in some cases is not recommended. I made some tests with our package collections and it worked as I expected. pkg-get is so fast solving dependencies at client side, but it depends on perl, so we should test all features and options to be sure. Also pkg-repgen needs a patch (where I'm working on) to avoid getting extra info (dependencies, README files and pre/post install scripts) from the prt-get's stuff installed at host. I think I could have this patch before the end of the week. So with this port I think we can tie up 2.6 development (core), then release RC7 and make a final test. If all goes well RC7 will become our final 2.6 release. At this point it would be good to publish a new tarball for each toolchain, which *could* be used to create more packages. I've forgotten something? Best regards, -- Jose V Beneyto | http://sepen.mine.nu/

2 1

RC6 available
by Victor Martinez 26 Aug '10

26 Aug '10

Hello, I just build the new 2 RC's, for eabi and noeabi, they can be found here. http://crux-arm.nu/releases/2.6/ At the moment I will let RC5 there too, until the new RC6 will be tested. I'm writing a little changelog about the RC's availables until this moment, with some notes I took while working on them. Best regards, Victor. Learning bit by bit Victor Martinez | http://lokalix.dyndns.org

2 3

pkgmk.in: sed patch to avoid development build references. Needs testing to confirm all files are clean.
by Victor Martinez 25 Aug '10

25 Aug '10

- find $PKG -type f -name '*.la' -exec sed -e "s|$CLFS||g" -e "s|$CROSTOOLS/$CTARGET|/usr|g" -i {} \; + find $PKG -type f -name '*.la' -exec sed -e "s|$CLFS||g" -e "s|$CROSTOOLS/$CTARGET|/usr|g" -e "s|$CROSSTOOLS||g" -i {} \; } check_footprint() {

1 1

Updated some opt-cross ports
by Victor Martinez 08 Jul '10

08 Jul '10

Hello, I just updated some ports without pushing the packages to the server. I prefer to wait, because it's needed the rebuild of some of others depending in the ones updated (not all). I have the packages in the server but not in the pkg repo. The list of updated ports is: [ports/opt-cross.git/2.6] whois: updated to 5.0.5 [ports/opt-cross.git/2.6] alsa-lib: updated to 1.0.23 [ports/opt-cross.git/2.6] alsa-utils: updated to 1.0.23 [ports/opt-cross.git/2.6] libpcap: updated to 1.1.1 [ports/opt-cross.git/2.6] glib: updated to 2.24.0 [ports/opt-cross.git/2.6] wgetpaste: updated to 2.16 [ports/opt-cross.git/2.6] tcpdump: updated to 4.1.1 [ports/opt-cross.git/2.6] ntp: updated to 4.2.6p1 [ports/opt-cross.git/2.6] libxml2: updated to 2.7.7 [ports/opt-cross.git/2.6] fetchmail: updated to 6.3.17 [ports/opt-cross.git/2.6] htop: 0.8.3 [ports/opt-cross.git/2.6] libpng: updated to 1.2.44 [ports/opt-cross.git/2.6] libtiff: updated to 3.9.4 [ports/opt-cross.git/2.6] lua: updated to 5.1.4 Along today and this weeked I'll try to make the updates of opt-cross and if I'm able I'll start with other collections (the core collection can be a good one to sync but may be when the current problems in the RC5 will be solved to prepare the final release synced with the last CRUX 2.6 state, explained in the next paragraph). We need to think about the final CRUX-ARM release. I think can be a good choice to bump all ports to the version used in the last stage of CRUX 2.6 and then rebuild the packages from all collections to fix the new updates and let all collections in a stable state (CRUX 2.7 is near, and soon all branches will be "closed", then it could be good time to make a sync between CRUX and CRUX-ARM for 2.6 version) Waiting for comments. Regards, Victor. --- Learning bit by bit Victor Martinez | http://lokalix.dyndns.org

1 0

[PATCH] pkgmk.in: Fix to remove hardcoded framework path in .la files
by Victor Martinez 10 Jun '10

10 Jun '10

- find $PKG -type f -name '*.la' -exec sed "s|$CLFS||g" -i {} \; + find $PKG -type f -name '*.la' -exec sed -e "s|$CLFS||g" -e "s|$CROSTOOLS/$CTARGET|/usr|g" -i {} \; } check_footprint() {

3 4

[PATCH] coreutils: Patch to don't build man pages (provided in a new source).
by Victor Martinez 10 Jun '10

10 Jun '10

name=coreutils version=8.4 -release=3 +release=4 source=(http://ftp.gnu.org/pub/gnu/$name/$name-$version.tar.gz - coreutils-uname.patch) + coreutils-uname.patch \ + coreutils-8.4_manpages.tar.bz2) build() { cd $name-$version patch -p1 -i $SRC/coreutils-uname.patch - - for f in $(find man/ -type f -name '*.x' -exec basename {} .x \;); do - touch $f $f.1 - done + sed "s|SUBDIRS = lib src doc man po tests gnulib-tests|SUBDIRS = lib src tests gnulib-tests|g" -i Makefile.in cat > config.cache <<EOF ac_cv_search_clock_gettime=no @@ -50,5 +48,5 @@ EOF install -d $PKG/bin mv $PKG/usr/bin/{cat,chgrp,chmod,chown,cp,date,dd,df,echo,false,ln,ls,mkdir,sleep} $PKG/bin mv $PKG/usr/bin/{mknod,mv,pwd,readlink,rm,rmdir,stty,sync,touch,true,uname} $PKG/bin - rm -r $PKG/usr/share + mv $SRC/usr/man $PKG/usr/ }

3 2

Doubts about 'bsdtar -m'
by Tilman Sauerbeck 03 Jun '10

03 Jun '10

Hi, in pkgutils-cross.git, commit 3f0931c4294981d3e3edb8c8ba8e422a4dd3a442 you added the -m flag to the bsdtar call that's used to extract source archives. That flag causes bsdtar to ignore the mtimes contained in the archive. This causes build failures for eg the module-init-tools port: In module-init-tools-3.11.1.tar.bz2, aclocal.m4 is stored after configure. When extracting with -m, aclocal.m4 will have a newer mtime than configure. When you call make, it will realize the time difference and will regenerate the configure script. For me, the newly created configure script then fails horribly. I don't think you should need -m. Can you give an example of what problem it fixed? Regards, Tilman -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?

2 7

[PATCH] Don't try to build binutils documentation.
by Tilman Sauerbeck 20 May '10

20 May '10

This removes the dependency on texinfo. --- Makefile | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/Makefile b/Makefile index 9fe0dd9..fd4a5d1 100644 --- a/Makefile +++ b/Makefile @@ -104,6 +104,7 @@ $(WORK)/binutils-$(BINUTILS_VERSION): $(WORK)/binutils-$(BINUTILS_VERSION).tar.b tar -C $(WORK) -xvjf $(WORK)/binutils-$(BINUTILS_VERSION).tar.bz2 cd $(WORK)/binutils-$(BINUTILS_VERSION) && \ patch -p1 -i $(WORK)/binutils-$(BINUTILS_VERSION)-branch_update-5.patch + sed -i '/^SUBDIRS/s/doc//' $(WORK)/binutils-$(BINUTILS_VERSION)/*/Makefile.in touch $(WORK)/binutils-$(BINUTILS_VERSION) $(WORK)/build-binutils: $(WORK)/binutils-$(BINUTILS_VERSION) -- 1.7.1

1 0

Native compilations on handhelds.
by Victor Martinez 17 May '10

17 May '10

Hello, I have been doing some test with native compilations under handhelds (at the moment I'm trying on htc-prophet). I noticed what Fredick told me about the -pipe flag, which seems to kill memory (only with 64Mb). I want to know your opinions about this and another topic related to native compilations and the commits in native git repos. I think it can be a good idea to remove the -pipe flags in ports which have it hardcoded (like perl). In this way we can let native building work with devices which don't have too much amount of ram. Related to ram are sources too, xz and lzma needs more ram to be uncompressed (their size is smaller that bz2 sources) and then it's impossible in devices with small amount of ram. I hope this can be interesting. Waiting your comments and suggestions. Regards, Victor. -- Learning bit by bit. Victor Martinez | http://lokalix.dyndns.org

3 5

Preparing the release CRUX-ARM 2.6
by pitillo＠crux-arm.nu 17 May '10

17 May '10

Hello, thinking in the first release of CRUX-ARM 2.6. What do you think to make it for next week? The 4th CRUX-ARM release candidate (RC4) has been tested for both, no-eabi and eabi. The current status for them is quite good: no-eabi: It's working fine on Jornada 720 (with a 2GB compact flash card, at the moment there are some problems[1]) and versatile (qemu) eabi: It's working fine on ipaq 191x, htc-prophet (qtek S200), htc-elf (htc Touch) and versatile (qemu). Suggestions, comments and opinions are welcome. Regards, Victor. [1] There are problems trying to run it in a 4GB compact flash. Working on progress about this topic.

2 3