lkml.org 
[lkml]   [2002]   [Feb]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH] 2.4.17 panic booting motherboards with hpt372
Date
A repost, yesterday I had the diff -u args reversed. Analysis is the same.

I have a kr7a-raid (hpt372 raid) motherboard.i The 2.4.17 kernel
(plus the 2.4.xx variants on the redhat 7.2 and mandrake 8.1)
panic during boot.

I have found the problem is in ide-pci.c (and a similar one in
hpt366.c) and diff -u follows below.

The code of interest begins at line 835 (2.4.17 base):
The hpt372 returns a class_rev of 5 which is not expected
by the switch statement.

pci_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev);
class_rev &= 0xff;

strcpy(d->name, chipset_names[class_rev]);

switch(class_rev) {
case 4:
case 3: printk("%s: IDE controller on PCI bus %02x dev
%02x\n",d->name, dev->bus->number, dev->devfn);
ide_setup_pci_device(dev, d);
return;
default: break;
}

The patch makes this code more defensive by using the highest known
class_rev if one is returned which is higher.

Also, for class_rev == 4, the strcpy copies a 7 byte string over
a 6 byte one ("HPT370A" over "HPT366") so I added a strncpy
to make this more defensive as well.

Roger Massey

--- drivers/ide/ide-pci.orig.c Mon Feb 4 19:37:50 2002
+++ drivers/ide/ide-pci.c Mon Feb 4 19:44:02 2002
@@ -836,7 +836,11 @@
pci_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev);
class_rev &= 0xff;

- strcpy(d->name, chipset_names[class_rev]);
+ if(class_rev >= (sizeof(chipset_names)/sizeof(char *))) {
+ class_rev = (sizeof(chipset_names)/sizeof(char *)) - 1;
+ }
+
+ strncpy(d->name, chipset_names[class_rev], strlen(d->name));

switch(class_rev) {
case 4:
--- drivers/ide/hpt366.orig.c Mon Feb 4 19:33:30 2002
+++ drivers/ide/hpt366.c Mon Feb 4 19:32:45 2002
@@ -214,6 +214,9 @@
pci_read_config_dword(bmide_dev, PCI_CLASS_REVISION, &class_rev);
class_rev &= 0xff;

+ if(class_rev >= (sizeof(chipset_names)/sizeof(char *)))
+ class_rev = (sizeof(chipset_names)/sizeof(char *)) -1;
+
/*
* at that point bibma+0x2 et bibma+0xa are byte registers
* to investigate:

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:24    [W:0.043 / U:0.180 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site