Re: [PATCH] fs/ceph/io: make ceph_start_io_*() killable

[Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>]

On Fri, 2024-12-06 at 17:50 +0100, Max Kellermann wrote:
> This allows killing processes that wait for a lock when one process
> is
> stuck waiting for the Ceph server.  This is similar to the NFS commit
> 38a125b31504 ("fs/nfs/io: make nfs_start_io_*() killable").
> 
> Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
> ---
>  fs/ceph/file.c | 22 +++++++++++++---------
>  fs/ceph/io.c   | 44 +++++++++++++++++++++++++++++++++-----------
>  fs/ceph/io.h   |  8 +++++---
>  3 files changed, 51 insertions(+), 23 deletions(-)
> 
> 

<skipped>

>  
>  /**
> diff --git a/fs/ceph/io.h b/fs/ceph/io.h
> index fa594cd77348..08d58253f533 100644
> --- a/fs/ceph/io.h
> +++ b/fs/ceph/io.h
> @@ -2,11 +2,13 @@
>  #ifndef _FS_CEPH_IO_H
>  #define _FS_CEPH_IO_H
>  
> -void ceph_start_io_read(struct inode *inode);
> +#include <linux/compiler_attributes.h> // for __must_check

Do we really need this comment (for __must_check)? It looks like not
very informative. What do you think?

I am not completely sure that it really needs to request compiler to
check that return value is processed. Do we really need to enforce it?

Thanks,
Slava.

> +
> +__must_check int ceph_start_io_read(struct inode *inode);
>  void ceph_end_io_read(struct inode *inode);
> -void ceph_start_io_write(struct inode *inode);
> +__must_check int ceph_start_io_write(struct inode *inode);
>  void ceph_end_io_write(struct inode *inode);
> -void ceph_start_io_direct(struct inode *inode);
> +__must_check int ceph_start_io_direct(struct inode *inode);
>  void ceph_end_io_direct(struct inode *inode);
>  
>  #endif /* FS_CEPH_IO_H */