Re: [PATCH 4/7] xfs: Implement fallocate query support mode

Re: [PATCH 4/7] xfs: Implement fallocate query support mode

[Christoph Hellwig]

This seems to miss any explanation and 6 out of 7 patches.

That being said, I really detest adding weird query modes to random
syscalls.  We should expose capabilities through the interface that
was designed for such things: (f)pathconf, as there are a lot more
important parameters we currently fail to expose to userspace to
start with.

Re: [PATCH 4/7] xfs: Implement fallocate query support mode

[OGAWA Hirofumi]

Christoph Hellwig <hch@infradead.org> writes:

I myself

Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>

to [5/7] fat part though,

> That being said, I really detest adding weird query modes to random
> syscalls.  We should expose capabilities through the interface that
> was designed for such things: (f)pathconf, as there are a lot more
> important parameters we currently fail to expose to userspace to
> start with.

This is true. For example, max link count. And it might be able to use
for fs's bug/issue detection too (in past, userland has to add
workaround for all cases, because noway to detect the buggy or not).

Thanks.
-- 
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>

Re: [PATCH 4/7] xfs: Implement fallocate query support mode

[Lukas Czerner]

On Mon, Sep 18, 2017 at 10:56:51AM -0700, Christoph Hellwig wrote:
> This seems to miss any explanation and 6 out of 7 patches.

See linux-fsdevel for entire patch set.
https://www.spinics.net/lists/linux-fsdevel/msg115740.html

> 
> That being said, I really detest adding weird query modes to random
> syscalls.  We should expose capabilities through the interface that
> was designed for such things: (f)pathconf, as there are a lot more
> important parameters we currently fail to expose to userspace to
> start with.

(f)pathconf on it's own is useless because I need to get this
information directly from the file system. Supported fallocate modes can
differ with inode flags (encrypted files, extent/indirect based files on
ext4). I also need in-kernel interface.

That's why I though fallocate is the convenient place to put this. I
have not looked to it, but other option might be to use fstat to get
this.

-Lukas

Re: [PATCH 4/7] xfs: Implement fallocate query support mode

[Christoph Hellwig]

On Tue, Sep 19, 2017 at 10:15:38AM +0200, Lukas Czerner wrote:
> On Mon, Sep 18, 2017 at 10:56:51AM -0700, Christoph Hellwig wrote:
> > This seems to miss any explanation and 6 out of 7 patches.
> 
> See linux-fsdevel for entire patch set.
> https://www.spinics.net/lists/linux-fsdevel/msg115740.html

But only one made it to linux-xfs, so please fix your patch sending.

> (f)pathconf on it's own is useless because I need to get this
> information directly from the file system. Supported fallocate modes can
> differ with inode flags (encrypted files, extent/indirect based files on
> ext4). I also need in-kernel interface.

And that is exactly what I want you to do:  finally bite the bullet
and add pathconf inode operation.