This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
pread, pwrite linux implementation does not conform to Single UnixSpecification
- From: "Gerhard Tonn" <TON at de dot ibm dot com>
- To: libc-alpha at sources dot redhat dot com
- Date: Wed, 17 Jul 2002 17:24:55 +0200
- Subject: pread, pwrite linux implementation does not conform to Single UnixSpecification
- Sensitivity:
Anybody any opinion on this subject? Is the implementation supposed to be
changed or does it work as designed? At least powerpc linux seems to have
made it SUS compliant in their special implementation.
Regards / Mit freundlichen Gruessen
Gerhard
Gerhard Tonn, Linux for eServer Development, +(49)-7031-16-4716, Lotus
Notes: ton@ibmde,
Internet: ton@de.ibm.com
---------------------- Forwarded by Gerhard Tonn/Germany/IBM on 17.07.2002
17:22 ---------------------------
Gerhard Tonn
10.07.2002 12:59
To: libc-alpha@sources.redhat.com
cc:
From: Gerhard Tonn/Germany/IBM@IBMDE
Subject: pread, pwrite linux implementation does not conform to Single
Unix Specification
Hi,
we have recently experienced problems with the ltp (linux test project)
testsuite, when passing a negative value as offset to pread and pwrite. The
functions don't return EINVAL as specified by
http://www.opengroup.org/onlinepubs/007908799/xsh/write.html and
http://www.opengroup.org/onlinepubs/007908799/xsh/read.html respectively,
but return EFBIG or no error at all. This problem appears on all platforms
using the sysdeps/unix/sysv/linux/ default implementation. The
implementation doesn't check for negative values and use the
__LONG_LONG_PAIR macro to convert from an off_t to a loff_t, which doesn't
do any sign extension. Could that be fixed?
Regards / Mit freundlichen Gruessen
Gerhard
Gerhard Tonn, Linux for eServer Development, +(49)-7031-16-4716, Lotus
Notes: ton@ibmde,
Internet: ton@de.ibm.com