Messages in this thread | | | Date | Thu, 8 Oct 1998 21:19:48 +0200 (MET DST) | From | Gerard Roudier <> | Subject | Re: cat /proc/pci and NCR 810 SCSI parity error |
| |
On Thu, 8 Oct 1998, Martin Mares wrote:
> Hello, > > > I'm not sure about what the PCI spec says about side effects. Having read > > side effect for something called a configuration space looks weird but > > would perhaps be acceptable. > > I don't have the PCI specs at hand now, but as far as I remember, side > effects on configuration space reads are explicitly forbidden.
Seems to me rather _implicitely_ assumed.
> > However the worst case is that some chips manage to lockup the whole PCI > > bus (and hence the system) when reading from some configuration registers. > > This is a clear violation of the PCI spec and a very serious one. > > And let's point our fingers at the culprit: it's Intel. *Again* (PIIX4 ACPI).
Culprit of what? Intel and Symbios break your preferred code by violating an implicitely assumed PCI requirement, but your preferred code may break running systems.
If I were in your shoes, I would not expect all users who will get their running system broken by 'lspci' to point their fingers in the direction you suggest. ;-)
> > Yes, but this type of things is unlikely to require frequent accesses to > > configuration space registers. This space is designed to be essentially > > safe to access, although maybe slow and I personnally hate side effects in > > this area. I prefer to be able to type lspci -xx when looking for PCI > > problems or debugging a driver (although in the last case I rather use > > lspci -xx -s device). > > I completely agree. /proc/bus/pci restricts non-root accesses to the > standard header (i.e., 64 bytes or 128 for CardBus bridges). Root can > read/write everything and there is no reason for disabling this capability -- > /proc/ioports is a clean precedent for root crashing the system by just reading > a file.
Are you requiring all Linux System administrator to be as knowledgeable as us on PCI technology? My opinion is that there is at least reason to _clearly_ discourage users to press this _red_ _button_, unless you just want _millions_ of fingers to point at your direction. ;-)
Regards, Gerard.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |