Re: Reworking suspend-resume sequence (was: Re: PCI PM: Restorestandard config registers of all devices early)

[Ingo Molnar]

* Ingo Molnar <mingo@xxxxxxx> wrote:

>  - the screaming-irq observation i had - do you consider that valid?:
> 
>    >> [ In theory this also solves screaming level-triggered irqs that 
>    >>   advertise themselves as edge-triggered [due to firmware/BIOS bug - 
>    >>   these do happen] and then keep spamming the system. ]
> 
>    I wanted to have a pretty much interchangeable flow method between edge 
>    and level triggered - so that the BIOS cannot screw us by enumerating an 
>    irq as edge-triggered while it's level-triggered.
> 
>    Especially for legacy x86 irqs in the low <16 range the trigger mode can 
>    be influenced by chipset settings and might not always be what we think 
>    it is.

For apic and MSI based methods that's not a big issue: the trigger mode is 
explicitly set by us, so if there's a mismatch it's a kernel bug.

And even for legacy ISA i8259a.c it should be fine after all, as we 
initialize it via:

        set_irq_chip_and_handler_name(irq, &i8259A_chip, handle_level_irq,

Which is a screaming-safe sequence. (and the x86 i8259 PIC does not lose 
edges.)

That still leaves other architectures ... but i think the argument is a lot 
weaker than i thought it to be.

Could you send a Signed-off-by so that i can queue it up and test it a bit? 
If it does not blow up on x86 in practice then we should be fine, and it 
avoids the MSI ->mask() stupidity as well.

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