[corinna-cygwin@cygwin.com: Re: ENOTSOCK errors with cygwin dll 1.3.21 and 1.3.22]

Corinna Vinschen vinschen@redhat.com
Mon May 26 15:31:00 GMT 2003

On Mon, May 26, 2003 at 11:05:49AM -0400, Pierre A. Humblet wrote:
> 1) The way I think you have it, a file can be made sparse even when 
>    O_CREAT | O_TRUNC isn't set. Is that desirable? 

Yes, I think so.  Why shouldn't a file become sparse when the accessing
application are going to create holes by seeking beyond EOF?   That's
the only situation in which this might happen.

> 2) in fhandler::write, test wincap.has_lseek_bug before the sparse file.
>    This is to cover the (hypothetical, so far) case where a win 95 would
> mount a 
>    remote file system supporting sparse files. I don't know if fs_flags 
>    would show the file can be sparse, nor if the win95 lseek bug would
> actually 
>    show up. So it may be paranoia.

You can test this, can't you?

> 4) I did some experiments on Win2000 with 4k clusters. The hole had to be
>    > 128kB before groups of 16 clusters would be skipped. Have you observed 
>    skipping at 64kB? 

I found that 128K clusters are used but you can write in the middle of
the hole and NTFS doesn't allocate more space then if it still fits
into the 128K cluster.  So the holes might become smaller than the
initial 128K.  Why not use 64K as lower border?  It doesn't hurt, IMO.

> 5) Not related, there is a bug on line 209 of fhandler.cc. test for size, 
>    not upper1.

You're going to send a patch, I assume.


Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Developer                                mailto:cygwin@cygwin.com
Red Hat, Inc.

More information about the Cygwin-developers mailing list