PATCH: handleIPError and need to import logging

[HARA Hiroshi]

Hi,

Here is the patch for "Add iSCSI target" error handling.
(anaconda does not handle Some incorrect situation well)

Would you review my patch, please?

thank you.

[David Cantrell]

On Thu, 21 Jun 2007 21:01:31 +0900
HARA Hiroshi  wrote:
> Hi,
>
> Here is the patch for "Add iSCSI target" error handling.
> (anaconda does not handle Some incorrect situation well)
>
> Would you review my patch, please?
>
> thank you.

The patch looks good.  I'll get it applied to rawhide.

Thanks,

Wait longer for usb floppy access (#211222)

[HARA Hiroshi]

Hi, Peter Jones and all.

I found the following changelog in RHEL5 anaconda.spec.

* Thu Jan  4 2007 Peter Jones  - 11.1.2.27-1
- Include cdroms in the scsi disk count (katzj, #207336)
- Translation display fixes (katzj, #216067)
- Wait longer for usb floppy access (#211222)   <-- here
- Make the package repo path list right (#221260)

and also find the following code in loader2/kickstart.c

/* usb can take some time to settle, even with the various hacks we
 * have in place.  so if we've been requested to ks from a floppy, try to
 * make sure there really isn't one before bailing */
while (!devices && (i++ < 10)) {
    logMessage(DEBUGLVL, "sleeping to wait for a floppy...");
    sleep(1);
    devices = probeDevices(CLASS_FLOPPY, BUS_MISC | BUS_IDE | BUS_SCSI,
PROBE_LOADED);
}

but sometimes anaconda can not do kickstart with usb-floppy and
go to reboot using anaconda-11.1.2.36-1(RHEL5 anaconda).
I found that even if devices is not NULL, it is possible that
devices[i]->device is NULL.

so I have three questions about it.
If you could (have a time), would you check these question, please?

1. in this case, do you think it is better that
   kudzu(not anaconda) care about it?
2. Would you give me your review about the attached patch?
3. How about CVS anaconda ?
   (CVS anaconda will not encounter this issue?)

thank you in advance

Wait longer for usb floppy access (#211222)

[HARA Hiroshi]

Hi, Peter Jones and all.

(maybe I sent the previous mail with wrong way,so I'll
resend this mail. if you already have read the previous one,
please ignore this.)

I found the following changelog in RHEL5 anaconda.spec.

* Thu Jan  4 2007 Peter Jones  - 11.1.2.27-1
- Include cdroms in the scsi disk count (katzj, #207336)
- Translation display fixes (katzj, #216067)
- Wait longer for usb floppy access (#211222)   <-- here
- Make the package repo path list right (#221260)

and also find the following code in loader2/kickstart.c

/* usb can take some time to settle, even with the various hacks we
 * have in place.  so if we've been requested to ks from a floppy, try to
 * make sure there really isn't one before bailing */
while (!devices && (i++ < 10)) {
    logMessage(DEBUGLVL, "sleeping to wait for a floppy...");
    sleep(1);
    devices = probeDevices(CLASS_FLOPPY, BUS_MISC | BUS_IDE | BUS_SCSI,
PROBE_LOADED);
}

but sometimes anaconda can not do kickstart with usb-floppy and
go to reboot using anaconda-11.1.2.36-1(RHEL5 anaconda).
I found that even if devices is not NULL, it is possible that
devices[i]->device is NULL.

so I have three questions about it.
If you could (have a time), would you check these question, please?

1. in this case, do you think it is better that
   kudzu(not anaconda) care about it?
2. Would you give me your review about the attached patch?
3. How about CVS anaconda ?
   (CVS anaconda will not encounter this issue?)

thank you in advance