This is the mail archive of the
cygwin@sources.redhat.com
mailing list for the Cygwin project.
Re: lseek() misbehaving.
- To: Jesper Eskilson <jojo at virtutech dot se>
- Subject: Re: lseek() misbehaving.
- From: Egor Duda <deo at logos-m dot ru>
- Date: Wed, 28 Feb 2001 18:17:05 +0300
- CC: cygwin at cygwin dot com
- Organization: DEO
- References: <u67ae7626sn.fsf@rachel.hq.vtech>
- Reply-To: Egor Duda <cygwin at cygwin dot com>
Hi!
Wednesday, 28 February, 2001 Jesper Eskilson jojo@virtutech.se wrote:
JE> read() read 32 bytes (requested 32) from fd 3
JE> =======
JE> b8 c0 7 8e d8 b8 0 90 8e c0 b9 0 1 29 f6 29 ff fc f3 a5 ea 19 0 0 90 bf f4 3f 8e d8 8e d0
JE> =======
...
JE> read() read 32 bytes (requested 32) from fd 3
JE> =======
JE> 89 fc 8e e1 bb 78 0 1e 64 c5 37 b1 6 fc 57 f3 a5 5f 1f c6 45 4 24 64 89 3f 64 8c 47 2 30 e4
JE> =======
JE> Now, you'd expect that the two 32-bytes block (32 is arbitrarily chosen)
JE> would be identical, but they are not. The (second) call to lseek() does not
JE> move the file-pointer at all, so the second 32-byte block is what you
JE> would've gotten if you read 64 bytes the first time.
i've checked in the patch for fhandler_dev_floppy::lseek several
hours ago. your program behaves correctly with my current cvs build
of cygwin1.dll
you can wait for a new snapshot, or build cygwin1.dll from current cvs
yourself, and check if it solves your problem.
Egor. mailto:deo@logos-m.ru ICQ 5165414 FidoNet 2:5020/496.19
--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple