Hello, I just wrote:
Hi there, Looks like your commit 74811f355f4f69a187fa74892dcf2a684b84ce99
(hpt366: convert to use ->host_priv)
Thanks for doing the bisection.
has caused the following crash:
Oops, we did it again. :-)
Already seeing Bart's mistake, see below:
(Full dmesg follows, as captured from a serial console)
[ 22.555200] hpt366: HPT374 chipset detected
[ 22.559489] hpt366 0000:03:06.0: IDE controller (0x1103:0x0008 rev 0x07)
[ 22.566278] HPT366_IDE 0000:03:06.0: PCI INT A -> GSI 28 (level, low) -> IRQ 28
[ 22.573729] usb 1-2: new full speed USB device using uhci_hcd and address 4
[ 22.580726] pci 0000:03:06.1: PCI INT A -> GSI 28 (level, low) -> IRQ 28
[ 22.587529] hpt366 0000:03:06.0: DPLL base: 48 MHz, f_CNT: 142, assuming 33 MHz PCI
[ 22.602705] hpt366 0000:03:06.0: using 50 MHz DPLL clock
[ 22.608181] hpt366 0000:03:06.0: 100% native mode on irq 28
[ 22.613831] hpt366 0000:03:06.1: no clock data saved by BIOS
Aha, HPT374 workaround for reading BIOS clock data didn't get executed.
[ 22.731015] usb 1-2: configuration #1 chosen from 1 choice
[ 22.737191] ata2.00: ATA-6: ST3120022A, 3.04, max UDMA/100
[ 22.738880] hub 1-2:1.0: USB hub found
[ 22.746597] ata2.00: 234441648 sectors, multi 16: LBA48
[ 22.747804] hpt366 0000:03:06.1: DPLL base: 33 MHz, f_CNT: 139, assuming
Uh, I wonder where did it get those 33 MHz DPLL base -- there's
simply no such base. Looks like memory addressed by .host_priv got
corrupt...
No, that's just wrong pointer arithmetics there:
struct hpt_info *info = host->host_priv + (hwif->dev == host->dev[1]);
The 'host_priv' is declared as 'void *', hence in case the condition abobe is
true, there will be added only 1 to host->host_priv, not sizeof(struct
hpt_info) as it should've been...
MBR, Sergei
--
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
