Re: [HTCondor-users] GPUs_MONITOR resource usage

[Michael Pelletier <Michael.V.Pelletier@xxxxxxxxxxxx>]

>  39637 ioctl(0, _IOC(_IOC_READ|_IOC_WRITE, 0x46, 0x2a, 0x20),

File descriptor 0 should be stdin, right? Odd that they've got a "write" associated with it. The second argument appears to be an expansion of this macro from ioctl.h:

#define _IOWR(type,nr,size) _IOC(_IOC_READ|_IOC_WRITE,(type),(nr),(_IOC_TYPECHECK(size)))

The type is 'F' from the ioctl-numbers.txt file, if I'm understanding this correctly:

https://github.com/torvalds/linux/blob/master/Documentation/ioctl/ioctl-number.txt

... which as you would expect is associated with video:

'F'	20	drivers/video/fsl-diu-fb.h	conflict!
'F'	20	drivers/video/intelfb/intelfb.h	conflict!
'F'	20	linux/ivtvfb.h		conflict!
'F'	20	linux/matroxfb.h	conflict!
'F'	20	drivers/video/aty/atyfb_base.c	conflict!
'F'	00-0F	video/da8xx-fb.h	conflict!
'F'	80-8F	linux/arcfb.h		conflict!
'F'          DD         video/sstfb.h conflict!

... and the "nr" value is dec42 (heh) which is a sequence number, and a "long double" type is a native one which has a size of 16 bytes.

Is the nouveau driver being used, or the NVIDIA-supplied driver?

Michael V. Pelletier
Information Technology
Digital Transformation & Innovation
Integrated Defense Systems
Raytheon Company