[K42-discussion] AIM7 runs, has problem removing SHM IDs with shmctl()

[Andrew Baumann]

Hi all,

I have AIM7 running and outputting numbers, but there's a bunch of warnings 
coming from a clean-up function which frees SHM regions, and I'm worried that 
despite the noise the inability to remove these regions may be affecting the 
performance.

I've attached a quick cut-and-paste of the relevant code as a test program. 
The relevant call is:
	shmctl(shmid, IPC_RMID, 0);
... which seems to get through to kitch-linux/lib/emu/shmctl.C, where the 
IPC_RMID case is stubbed out to EINVAL. It looks like Maria's code... any 
pointers on how tricky doing the removal would be? Or even if it is a problem 
I should be worring about -- I guess for now I could just comment out the 
warning and continue.

I'll next have a look at the SDET scripts and see whether I can run AIM7 the 
same way. Once I get it going, I'll write up a HOWTO-type document for the 
wiki. Thanks again for the help with this so far.

Andrew
-------------- next part --------------
A non-text attachment was scrubbed...
Name: shmtest.c
Type: text/x-csrc
Size: 4680 bytes
Desc: not available
Url : http://ozlabs.org/pipermail/k42-discussion/attachments/20050919/edf480b7/attachment.c