Well, I can confirm that it is in the errata internally available in a
document marked 5/14/2001. I had difficulty finding any errata
listings for the 7400 on our external site, though.
> There's more than just a process difference though.
> The version number is seriously different. It's not
> just one bit changing to indicate a different process.
> Here I am, with a version 2.9 chip:
> cpu : 7400, altivec supported
> temperature : 35-40 C (uncalibrated)
> clock : 450MHz
> revision : 2.9 (pvr 000c 0209)
Yeah, the 7410 had some small microarchitectural and architectural
changes. Changes that mean it can't necessarily be used in the same
socket as the 7400. Votages were modified, and some memory interfacing
was changed. Also, it was a significantly different process (a major
shrink), so it gets a new name. The various versions of the 7400 are
all in the same sized process (with slightly different transistors, I
> Any one of the counters would do; the event can be
> moved around as needed. Also note the TBSEL bits in
> MMCR0. TBSEL gives another way to get an interrupt,
> without giving up any of the counters.
True. But I'm going to have to think about this one more, before I
agree with you. :)
> Pardon me for being a pessimist. I have to imagine
> that the counters don't turn off fast enough too.
I'm fairly certain they do shut off the instant the interrupt happens
if they aren't supposed to count privileged events, but I'd have to do
some testing to prove it. The issue is that MSR[EE] takes 2 cycles to
>> Is this bug restricted to 7400/7410 only, or does it
>> affect the 750 (and relatives) and 604/604e too?
>> I'm thinking about ppc support for my perfctr driver,
>> and whether overflow interrupts are worth supporting
>> or not given the errata.
> 604/604e doesn't even have performance monitoring AFAIK.
> I've heard nothing to suggest that the 750 is affected.
> I'll give you a hand; point me to the latest perfctr code
> and explain how it is supposed to interact with oprofile.
The 604 and 604e both have performance monitors. I can't find anything
which says that this is a bug in them. However, being as this bug
managed to creep into the 7400/7410, I'm not willing to say that the
bug didn't exist in all of those processors. More testing would be
BTW, I am also interested in helping out with this code. Meanwhile,
I'll attempt to find the answers to those questions.
PowerPC Software Enablement
Note that my opinions are not Motorola's, even the good ones!
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to firstname.lastname@example.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/