[OpenAFS-port-darwin] fs listaliases broken?

[David Botsch]

Using ktrace, I see the following when doing a fs listal:

 1129 fs       CALL  open(280448,2,-2146680062)
  1129 fs       NAMI  "/dev/openafs_ioctl"
  1129 fs       RET   open 4
  1129 fs       CALL  ioctl(4,-1071627519,-1073743696)
  1129 fs       RET   ioctl 0
  1129 fs       CALL  close(4)
  1129 fs       RET   close 0
  1129 fs       CALL  fstat(1,-1073745272)
  1129 fs       RET   fstat 0
  1129 fs       CALL  ioctl(1,1074030202,-1073745400)
  1129 fs       RET   ioctl 0
  1129 fs       CALL  write(1,442368,35)
  1129 fs       GIO   fd 1 wrote 35 bytes
       "Alias msc for cell msc.cornell.edu
       "
  1129 fs       RET   write 35/0x23
  1129 fs       CALL  open(280448,2,-2146680062)
  1129 fs       NAMI  "/dev/openafs_ioctl"
  1129 fs       RET   open 4
  1129 fs       CALL  ioctl(4,-1071627519,-1073743696)
  1129 fs       RET   ioctl -1 errno 33 Numerical argument out of domain
  1129 fs       CALL  close(4)
  1129 fs       RET   close 0
 1129 fs       CALL  write(1,442368,17)
  1129 fs       GIO   fd 1 wrote 17 bytes
       "Alias  for cell 
       "
  1129 fs       RET   write 17/0x11
  1129 fs       CALL  open(280448,2,-2146680062)
  1129 fs       NAMI  "/dev/openafs_ioctl"
  1129 fs       RET   open 4
  1129 fs       CALL  ioctl(4,-1071627519,-1073743696)
  1129 fs       RET   ioctl -1 errno 33 Numerical argument out of domain
  1129 fs       CALL  close(4)
  1129 fs       RET   close 0
  1129 fs       CALL  write(1,442368,17)
  1129 fs       GIO   fd 1 wrote 17 bytes
       "Alias  for cell 
       "
  1129 fs       RET   write 17/0x11
  1129 fs       CALL  open(280448,2,-2146680062)
  1129 fs       NAMI  "/dev/openafs_ioctl"
  1129 fs       RET   open 4
  1129 fs       CALL  ioctl(4,-1071627519,-1073743696)
  1129 fs       RET   ioctl -1 errno 33 Numerical argument out of domain
 1129 fs       CALL  close(4)
  1129 fs       RET   close 0
  1129 fs       CALL  write(1,442368,17)
  1129 fs       GIO   fd 1 wrote 17 bytes
       "Alias  for cell 
       "

So, the code for fs.c only breaks out of the loop if the ioctl errno returned
is EDOM, and we are getting errno 33, which IS EDOM. So, this would seem
to be correct. However, doing a printf, I get:

 ./fs listal |head
Code is 0 w errno 2
Alias msc for cell msc.cornell.edu
Code is 78 w errno 33
Alias  for cell
Code is 78 w errno 33
Alias  for cell
Code is 78 w errno 33
Alias  for cell
Code is 78 w errno 33
Alias  for cell 

So, the value of code is not being set to what the ioctl is returning.
Interesting...

On Tue, Feb 14, 2006 at 04:20:34PM -0500, Chaskiel M Grundman wrote:
> --On Monday, February 13, 2006 23:10:10 -0500 David Botsch 
> <dwb7@ccmr.cornell.edu> wrote:
> 
> >Related perhaps to the fs listcells bug I reported?
> >
> >On Mon, Feb 13, 2006 at 10:10:39PM -0500, Gedaliah Wolosh wrote:
> >>
> >>loops until I hit ^C
> >>
> >>Alias  for cell
> 
> 
> I can reproduce this on macos, but not other platforms running 1.4.1. 
> there's no macos-specific code in the listcells pioctl.... I have not yet 
> figured out what's actually going on.
> 
> _______________________________________________
> port-darwin mailing list
> port-darwin@openafs.org
> https://lists.openafs.org/mailman/listinfo/port-darwin

-- 
********************************
David William Botsch
Consultant/Advisor II
CCMR Computing Facility
dwb7@ccmr.cornell.edu
********************************