Re: [PATCH v4 2/2] gpu: nova-core: add support for GPU timer block. (v4)

["Alexandre Courbot" <acourbot@nvidia.com>]

On Fri Nov 6, 2026 at 11:41 PM JST, Alexandre Courbot wrote:
<...>
>> nsec_lo/hi are now used in the register from_nsec for both tu102 and
>> gh100 registers.
>>
>>> I'd ignore Sashiko's suggestion as it looks like it is trying to
>>> reconcile the code against the comment instead of the other way around.
>>
>> I just fixed things to truncate and used wrapping_sub, to be honest
>> being careful here is probably the wrong answer, my local review bot
>> pointed out some wierdness can occur in 2043 with the current scheme,
>> so let's just keep it simple, wrapping subtract, and feed the result
>> to the hardware which is what the other driver does and what nouveau
>> did.
>
> The other drivers do what they do, but here again do we understand what
> happens when we write a wrapped value into the timer registers?

I checked what OpenRM does and it doesn't appear to wrap. Actually
`tmrSetCurrentTime_GH100` has the following line:

    NV_ASSERT_OR_RETURN(secTimerNs < osTimeNs, NV_ERR_INVALID_STATE);

meaning it expects the OS/GPU time delta to be positive. The error it
returns in that case is ignored by the caller, so nothing gets
programmed. In the absence of a better alternative I guess we should do
the same thing, i.e. the `checked_sub` and return early (but the
returned error should be just logged and ignored).

Nouveau AFAICT does not write `NV_PGC6_SCI_SYS_TIMER_OFFSET_*` at all.