Arch Linux User Repository

If there are bugs related to the PKGBUILD or config, report here.

Otherwise, please report here:

https://github.com/koverstreet/bcachefs/issues

Issue for the build failures regarding the missing symbols as reported in comments

• https://aur.archlinux.org/packages/linux-bcachefs-git#comment-931358
• https://aur.archlinux.org/packages/linux-bcachefs-git#comment-929632

reported upstream here: https://github.com/koverstreet/bcachefs/issues/590

@ticpu

Please report upstream if you haven't already.

Shouldn't have to work around stuff in config... but when even that doesn't work... with a filesystem have little knowledge about the inner workings of, I find it far safer to just leave things to Kent or the others working on bcachefs to resolve, and so present fixes for.

I guess I don't want to be responsible for creating weird bugs that might eat people's data, heh.

The latest version gives me the same issue as shintak but enabling CONFIG_LOCK_STAT didn't fix it. I had to remove the 2 tracepoints in the code to finish compiling.

diff --git a/fs/bcachefs/six.c b/fs/bcachefs/six.c
index 581aee565e95..4fce655a94e9 100644
--- a/fs/bcachefs/six.c
+++ b/fs/bcachefs/six.c
@@ -456,7 +456,6 @@ static int six_lock_slowpath(struct six_lock *lock, enum six_lock_type type,
                smp_mb__after_atomic();
        }

-       trace_contention_begin(lock, 0);
        lock_contended(&lock->dep_map, ip);

        if (six_optimistic_spin(lock, type))
@@ -526,7 +525,6 @@ static int six_lock_slowpath(struct six_lock *lock, enum six_lock_type type,
                six_clear_bitmask(lock, SIX_LOCK_HELD_write);
                six_lock_wakeup(lock, atomic_read(&lock->state), SIX_LOCK_read);
        }
-       trace_contention_end(lock, 0);

        return ret;
 }

@shintak

Please report this upstream as soon as possible, as it will help immensely with Kent fixing these sort of bugs.

Enabling CONFIG_LOCK_STAT bloats the install size quite a bit... not sure whether to actually enable it, or just wait until Kent fixes said bug.

Maybe just me but current version fails building with follwing error:

  MODPOST Module.symvers
ERROR: modpost: "__SCT__tp_func_contention_begin" [fs/bcachefs/bcachefs.ko] undefined!
ERROR: modpost: "__SCT__tp_func_contention_end" [fs/bcachefs/bcachefs.ko] undefined!
ERROR: modpost: "__tracepoint_contention_begin" [fs/bcachefs/bcachefs.ko] undefined!
ERROR: modpost: "__tracepoint_contention_end" [fs/bcachefs/bcachefs.ko] undefined!
ERROR: modpost: "__SCK__tp_func_contention_end" [fs/bcachefs/bcachefs.ko] undefined!
ERROR: modpost: "__SCK__tp_func_contention_begin" [fs/bcachefs/bcachefs.ko] undefined!
make[1]: *** [scripts/Makefile.modpost:136: Module.symvers] エラー 1
make: *** [Makefile:1978: modpost] エラー 2

Downloaded the package and after adding CONFIG_LOCK_STAT=y to config, build succeeded. However, I think it is more of a bug on the bcachefs side.

Thanks for the quick update :)

It's currently broken with fatal: reference is not a tree: be3ef67f38e1e6a6bdbe2253b6bf4c19be05dcb9

@QuartzDragon Thanks for quick fix. Confirmed now install succeeds.

@shintak

Fixed. The repo maintainer likes to force-push fixes a lot at the moment, so the hash changes. :/

Repo seemingly won't build beyond a certain commit at the moment, so I purposefully keep it back by default.

Tried installing, but it fails checking out specific commit:

Cloning into 'linux-bcachefs'...
done.
Updating files: 100% (80529/80529), done.
fatal: reference is not a tree: be3ef67f38e1e6a6bdbe2253b6bf4c19be05dcb9

Looks like this commit cant reach from any branch, so git clone doesn't download the commit object. https://github.com/koverstreet/bcachefs/commit/be3ef67f38e

This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.