Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cornell!uw-beaver!rice!sun-spots-request
From: operldl@griffin.uvm.edu (Louis)
Newsgroups: comp.sys.sun
Subject: sun os bug in semop() handling
Keywords: SunOS
Message-ID: <1168@uvm-gen.UUCP>
Date: 11 May 89 09:56:37 GMT
Sender: usenet@rice.edu
Organization: EMBA Computer Facility, Univ. of Vermont, Burlington.
Lines: 24
Approved: Sun-Spots@rice.edu
Original-Date: 3 May 89 17:05:47 GMT
X-Sun-Spots-Digest: Volume 7, Issue 282, message 19 of 19

The system call semop(), reportedly (according to man 2 semop) returns the
value of semval after the operation.  Unfortunately this isn't the case in
release 4.0.  What it does return is 0, or -1 on error, which is not
sufficient.

1. Has anyone else encountered this bug?

I have traced down the mistake to the system's semop() function. This
function unlike semctl() in the same file, does not set up u.u_rval1 =
semval.

The fix should be to set `u.u_rval1 = semval' in the source code. However
since I've never mucked with fixing the UNIX os before I'm unsure of the
best way to go about this.

I have checked the latest Sun bug reports & found no mention of this bug.

2. Some tips about this would also be helpful.

Please send any replies to me at:

			operldl@gen.uvm.edu

							Thanks.