<div dir="ltr">I was hoping to get this simple work around in all release branches and then if the root issue is found and fixed, we can test it a bit longer in master.<div><br></div><div>I'll double check the actual commit, but I'm sure I see this behavior in dizzy (and the work around was also first tested in dizzy, because that was blocking us in building small enough initramfs image).</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Apr 21, 2015 at 8:24 AM, Richard Purdie <span dir="ltr"><<a href="mailto:richard.purdie@linuxfoundation.org" target="_blank">richard.purdie@linuxfoundation.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Mon, 2015-04-20 at 15:05 +0200, Martin Jansa wrote:<br>
> On Fri, Apr 10, 2015 at 04:06:14PM +0200, Martin Jansa wrote:<br>
> > * install populate-extfs.sh from contrib, be aware that in order<br>
> > to use it you need to set DEBUGFS shell variable, otherwise it will<br>
> > try to use debugfs from relative path which is almost always<br>
> > incorrect:<br>
> > CONTRIB_DIR=$(dirname $(readlink -f $0))<br>
> > DEBUGFS="$CONTRIB_DIR/../debugfs/debugfs"<br>
> ><br>
> > * use symlinks to install e2fsck, mke2fs, without this option we have:<br>
> > $ ls -lahi tmp-glibc/work/arm920tt-oe-linux-gnueabi/e2fsprogs/1.42.9-r0/image/sbin/<br>
> > ...<br>
> > 101982430 -rwxr-xr-x 5 bitbake bitbake 972K Apr 10 15:43 e2fsck<br>
> > 101982430 -rwxr-xr-x 5 bitbake bitbake 972K Apr 10 15:43 fsck.ext2<br>
> > 101982430 -rwxr-xr-x 5 bitbake bitbake 972K Apr 10 15:43 fsck.ext3<br>
> > 101982430 -rwxr-xr-x 5 bitbake bitbake 972K Apr 10 15:43 fsck.ext4<br>
> > 101982430 -rwxr-xr-x 5 bitbake bitbake 972K Apr 10 15:43 fsck.ext4dev<br>
> > ...<br>
> > 101982441 -rwxr-xr-x 5 bitbake bitbake 348K Apr 10 15:43 mke2fs<br>
> > 101982441 -rwxr-xr-x 5 bitbake bitbake 348K Apr 10 15:43 mkfs.ext2<br>
> > 101982441 -rwxr-xr-x 5 bitbake bitbake 348K Apr 10 15:43 mkfs.ext3<br>
> > 101982441 -rwxr-xr-x 5 bitbake bitbake 348K Apr 10 15:43 mkfs.ext4<br>
> > 101982441 -rwxr-xr-x 5 bitbake bitbake 348K Apr 10 15:43 mkfs.ext4dev<br>
> > ...<br>
> ><br>
> > which would be OK, because they are hadlinks, but after runstrip in<br>
> > do_package we get one stripped binary and 4 hardlinks to the original<br>
> > one:<br>
> ><br>
> > $ ls -lahi tmp-glibc/work/arm920tt-oe-linux-gnueabi/e2fsprogs/1.42.9-r0/packages-split/e2fsprogs-e2fsck/sbin<br>
> > 101982713 -rwxr-xr-x 8 bitbake bitbake 972K Apr 10 15:43 e2fsck<br>
> > 101982713 -rwxr-xr-x 8 bitbake bitbake 972K Apr 10 15:43 fsck.ext2<br>
> > 101982713 -rwxr-xr-x 8 bitbake bitbake 972K Apr 10 15:43 fsck.ext3<br>
> > 101982713 -rwxr-xr-x 8 bitbake bitbake 972K Apr 10 15:43 fsck.ext4<br>
> > 101983136 -rwxr-xr-x 2 bitbake bitbake 185K Apr 10 15:43 fsck.ext4dev<br>
> > $ ls -lahi tmp-glibc/work/arm920tt-oe-linux-gnueabi/e2fsprogs/1.42.9-r0/packages-split/e2fsprogs-mke2fs/sbin/<br>
> > 101982716 -rwxr-xr-x 8 bitbake bitbake 348K Apr 10 15:43 mke2fs<br>
> > 101982716 -rwxr-xr-x 8 bitbake bitbake 348K Apr 10 15:43 mkfs.ext2<br>
> > 101988266 -rwxr-xr-x 2 bitbake bitbake 72K Apr 10 15:43 mkfs.ext3<br>
> > 101982716 -rwxr-xr-x 8 bitbake bitbake 348K Apr 10 15:43 mkfs.ext4<br>
> > 101982716 -rwxr-xr-x 8 bitbake bitbake 348K Apr 10 15:43 mkfs.ext4dev<br>
> ><br>
> > That's super annoying for big files like this which are often include<br>
> > in small recovery images. Using --enable-symlink-install option results<br>
> > in one stripped binary and 4 symlinks:<br>
> ><br>
> > $ ls -lahi tmp-glibc/work/arm920tt-oe-linux-gnueabi/e2fsprogs/1.42.9-r0/packages-split/e2fsprogs-e2fsck/sbin/<br>
> > 102113806 -rwxr-xr-x 2 bitbake bitbake 185K Apr 10 15:50 e2fsck<br>
> > 102113813 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 fsck.ext2 -> e2fsck<br>
> > 102113814 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 fsck.ext3 -> e2fsck<br>
> > 102113812 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 fsck.ext4 -> e2fsck<br>
> > 102113815 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 fsck.ext4dev -> e2fsck<br>
> > $ ls -lahi tmp-glibc/work/arm920tt-oe-linux-gnueabi/e2fsprogs/1.42.9-r0/packages-split/e2fsprogs-mke2fs/sbin/<br>
> > 102113804 -rwxr-xr-x 2 bitbake bitbake 72K Apr 10 15:50 mke2fs<br>
> > 102113825 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 mkfs.ext2 -> mke2fs<br>
> > 102113826 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 mkfs.ext3 -> mke2fs<br>
> > 102113823 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 mkfs.ext4 -> mke2fs<br>
> > 102113824 lrwxrwxrwx 1 bitbake bitbake 6 Apr 10 15:50 mkfs.ext4dev -> mke2fs<br>
> ><br>
> > Saving cca 1,5MB.<br>
><br>
> Ping, this is needed also in fido and dizzy branches.<br>
<br>
</div></div>I'm still hoping we can fix the packaging code so that we avoid this<br>
problem with hardlinked files which is why I'm holding off this<br>
slightly.<br>
<br>
Also, are you sure this problem applies in dizzy? The change the<br>
introduced this problem wasn't in dizzy as far as I remember but I could<br>
be wrong...<br>
<br>
Cheers,<br>
<br>
Richard<br>
<br>
<br>
</blockquote></div><br></div>