Matthew Wilcox wrote:
> On Thu, Mar 18, 2010 at 03:31:55PM -0400, Don Dutile wrote:
>> Matthew Wilcox wrote:
>>> So we need to figure out why your machine doesn't support extended
>>> config space. What chipset does it use? Is there a BIOS update available
>>> for it? What does it report in dmesg for config space accesses?
>> In addition to having MMCONFIG enabled on the system (hw & kernel),
>> you also need to be root because of this snippet in drivers/pci/pci-sysfs.c,
>> pci_read_config():
>> unsigned int size = 64;
>> loff_t init_off = off;
>> u8 *data = (u8*) buf;
>>
>> /* Several chips lock up trying to read undefined config space */
>> if (capable(CAP_SYS_ADMIN)) {
>> size = dev->cfg_size;
>> } else if (dev->hdr_type == PCI_HEADER_TYPE_CARDBUS) {
>> size = 128;
>> }
>>
>> which will truncate the read to 1st 64-bytes, so unpriv user can't
>> crash a system that has badly behaving PCI(e) config space.
>>
>> IMO, the CAP_SYS_ADMIN ought to be replaced w/a quirk check
>> for badly-behaving devices &/or chip sets.
>
> Joyce already said at the start of the thread that they were able to
> get 256 bytes of config space. Please don't hijack this thread where
> a user is receiving help.
>
My apologies. I thought I was helping. I wasn't trying to hijack the thread.
sorry for missing the '256' number in initial msg.
- Don
--
To unsubscribe from this list: send the line "unsubscribe linux-pci" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
