On 08/13/2011 06:57 PM, Peter Naulls wrote:
On 08/13/2011 05:31 PM, Peter Naulls wrote:
On 08/13/2011 02:45 PM, Peter Naulls wrote:
So, I don't have much to go on. Suggestions on how to resolve
this welcome.
From echo t > /proc/sysrq-trigger
This appears to be a recurrence of an old kernel bug. I reintroduced
some code that was present in e.g, 2.6.32.10 (you can see this mentioned
in the G300NH boot load on the wiki page). The following survived an
hour of writes before I stopped it.
--- linux-2.6.39.2/drivers/mtd/chips/cfi_cmdset_0002.c.orig 2011-08-14
10:15:53.960554911 -0700
+++ linux-2.6.39.2/drivers/mtd/chips/cfi_cmdset_0002.c 2011-08-14
10:17:25.273007715 -0700
@@ -688,6 +688,8 @@
return 0;
case FL_ERASING:
+ if (mode == FL_WRITING) goto sleep;
+
if (!cfip || !(cfip->EraseSuspend & (0x1|0x2)) ||
!(mode == FL_READY || mode == FL_POINT ||
(mode == FL_WRITING && (cfip->EraseSuspend & 0x2))))
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
