Re: LFS build of File-5.45 problem stops on compilation of seccomp.c
"dave kohler" ([email protected] via lfs-support Mailing List) <[email protected]> Wed, 19 Feb 2025 18:46:43 -0600
| Newsgroups | gmane.linux.lfs.support |
|---|---|
| Message-ID | <CAE62XYo+cmdMwcULteHTdhQYXy2HvN-Vva5J8oNdD7vKxDKXDQ@mail.gmail.com> |
I have audited all the symbolic links that the LFS 12.3-RC1 book directed
me to make up to the point of stopping with the error. All the symbolic
links appear to be correctly made. I did this by navigating to the target
directory of each ln -s I found and performed yesterday, then used ls -alt
to confirm the link was correctly made. So I doubt the symbolic links are
part of the problem in my LFS build. Hope you can figure out the cause of
the problems with output of the modified gcc test command:
SEARCH_DIR("=/mnt/lfs/tools/x86_64-lfs-linux-gnu/lib64");
SEARCH_DIR("=/usr/local/lib64"); SEARCH_DIR("=/lib64"); SE
ARCH_DIR("=/usr/lib64");
SEARCH_DIR("=/mnt/lfs/tools/x86_64-lfs-linux-gnu/lib");
SEARCH_DIR("=/usr/local/lib"); SEAR
CH_DIR("=/lib"); SEARCH_DIR("=/usr/lib");
iIf you have run out of ideas, please let me know.
Then I will purge the partitions on the lfs drive and reformat, re-download
the source tarballs, then try and rebuild everything per the RC1 book up
through the end of chapter 6. I will report back on this thread with the
results in that eventuality.
On Wed, Feb 19, 2025 at 4:28 PM dave kohler <[email protected]> wrote:
> Another thought. Could this be the result of a soft link being setup
> incorrectly. I did note that on my subsequent attempts to rebuild the LFS
> project from the start, I was being told that the soft link was already
> found. or something like that.
> I don't know much about soft links, especially how to purge them.
>
>
> On Wed, Feb 19, 2025 at 4:21 PM dave kohler <[email protected]> wrote:
>
>> When I rebuilt according to LFS 12.3-RC1 I had gained a lot of
>> experience on using the book's instructions. When I enter anything other
>> than simple one word commands, I learned to just copy/paste the multiline
>> commands from the web page to the terminal window. so a manual mistyping
>> of --with-sysroot=$LFS is pretty unlikely.
>>
>>
>> On Wed, Feb 19, 2025 at 4:14 PM dave kohler <[email protected]> wrote:
>>
>>> I copy/pasted your test into my terminal window and it outputs
>>> '/mnt/lfs'
>>>
>>> as I think it should.
>>>
>>> On Wed, Feb 19, 2025 at 4:03 PM Xi Ruoyao <
>>> [email protected]> wrote:
>>>
>>>> On Wed, 2025-02-19 at 15:58 -0600, dave kohler wrote:
>>>> > here is the output to the modified test gcc command:
>>>> >
>>>> > lfs:/mnt/lfs$ echo "int main(){}" | $LFS_TGT-gcc -x c - -Wl,--verbose
>>>>
>>>> /* snip */
>>>>
>>>> > SEARCH_DIR("=/mnt/lfs/tools/x86_64-lfs-linux-gnu/lib64");
>>>> SEARCH_DIR("=/usr/local/lib64"); SEARCH_DIR("=/lib64"); SE
>>>> > ARCH_DIR("=/usr/lib64");
>>>> SEARCH_DIR("=/mnt/lfs/tools/x86_64-lfs-linux-gnu/lib");
>>>> SEARCH_DIR("=/usr/local/lib"); SEAR
>>>> > CH_DIR("=/lib"); SEARCH_DIR("=/usr/lib");
>>>>
>>>> Now try "$LFS_TGT-ld -print-sysroot". It should output /mnt/lfs.
>>>>
>>>> If it outputs nothing instead, it means --with-sysroot=$LFS wasn't
>>>> correctly passed for binutils pass 1. Maybe $LFS wasn't correctly set,
>>>> or you've mistyped it (maybe something like --with-sysroot=$LFs).
>>>>
>>>> If my assumption is true you need to purge the broken LFS build and
>>>> start over (and making sure you've typed all the commands correctly this
>>>> time). Simply rebuilding binutils pass 1 and continuing will likely
>>>> cause a failure at the start of chapter 7.
>>>>
>>>> --
>>>> Xi Ruoyao <[email protected]>
>>>> School of Aerospace Science and Technology, Xidian University
>>>> --
>>>> http://lists.linuxfromscratch.org/sympa/info/lfs-support
>>>> Unsubscribe: See the above information page
>>>
>>>
--
http://lists.linuxfromscratch.org/sympa/info/lfs-support
Unsubscribe: See the above information page