UPDATE: ATA mkIII first official patches - please test!

[David O'Brien]

On Sun, Feb 13, 2005 at 07:08:33PM +0100, Anthony Ginepro wrote:
> - Retrying "call doadump" under a kernel without touching ATA_R_DEBUG,
> it worked eventually once and could be canceled. Most of the time, dump
> can't be canceled and doesn't progress.
> 
> Should I open a PR so that the issue can be tracked ?

This is what I have to do with ATAng to dump to an IDE disk, via Peter Wemm.
Maybe it will help you find a work around with mkIII.

Index: ata-all.c
===================================================================
RCS file: /home/ncvs/src/sys/dev/ata/ata-all.c,v
retrieving revision 1.234
diff -u -r1.234 ata-all.c
--- ata-all.c	24 Nov 2004 10:47:26 -0000	1.234
+++ ata-all.c	3 Dec 2004 00:46:03 -0000
@@ -375,6 +375,8 @@
     struct ata_channel *ch;
     int ctlr;
 
+if (panicstr) return;
+
     /* flush cache on all devices */
     for (ctlr = 0; ctlr < devclass_get_maxunit(ata_devclass); ctlr++) {
 	if (!(ch = devclass_get_softc(ata_devclass, ctlr)))