xfs
[Top] [All Lists]

Re: [PATCH 6/6] Add configure check for members of dirent for use in io/

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH 6/6] Add configure check for members of dirent for use in io/readdir.c
From: Felix Janda <felix.janda@xxxxxxxxx>
Date: Thu, 14 Jan 2016 20:06:34 +0100
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20160114101729.GC27644@xxxxxxxxxxxxx>
References: <cover.1452627740.git.felix.janda@xxxxxxxxx> <20160112200022.GG568@nyan> <20160113075855.GG21939@xxxxxxxxxxxxx> <20160113170938.GA10942@nyan> <20160114101729.GC27644@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.23 (2014-03-12)
Christoph Hellwig wrote:
> On Wed, Jan 13, 2016 at 06:09:38PM +0100, Felix Janda wrote:
> > Christoph Hellwig wrote:
> > > I think we're having a deeper issue here.  For an LFS build we
> > > absolutely should have these values as the kernel dirent64 has them.
> > > 
> > > Either we manage to not pull in the LFS version of readdir/getdents
> > > or musl is doing something inredibly stupid.  Can you investigate if
> > > and why we're not using the LFS getdents64 here?
> > 
> > With musl you always have LFS. However still musl does not define the
> > _DIRENT_HAVE_*. You can see the full dirent.h at:
> > 
> > http://git.musl-libc.org/cgit/musl/tree/include/dirent.h
> > 
> > I will ask on the musl list whether these constants can be added. So
> > you can disregard this patch.
> 
> Oh, ok.  That's a little weird, but I guess working around that
> with a new autoconf test is fine.

Actually, it should also be useful for other systems. For example, on
FreeBSD dirent has d_type (but does not declare _DIRENT_HAVE_D_TYPE),
but currently dump_dirent() does not print it.

Felix

<Prev in Thread] Current Thread [Next in Thread>