ProfSampleCostCentre in the eventlog file carries a CapNo instead of a Capset

[nfrisby]

I think the first field of ProfSampleCostCentre should be profCap :: !CapNo instead of profCapset :: !Capset.

ghc-events/src/GHC/RTS/EventTypes.hs

Lines 437 to 442 in 0fb2c17

	| ProfSampleCostCentre
	{ profCapset :: !Capset
	, profTicks :: !Word64
	, profStackDepth :: !Word8
	, profCcsStack :: !(VU.Vector Word32)
	}

ghc-events/src/GHC/RTS/Events/Binary.hs

Lines 895 to 914 in 85b9129

	timeProfParsers :: [EventParser EventInfo]
	timeProfParsers = [
	FixedSizeParser EVENT_PROF_BEGIN 8 $ do
	profTickInterval <- get
	return $! ProfBegin{..}
	, VariableSizeParser EVENT_PROF_SAMPLE_COST_CENTRE $ do
	payloadLen <- get :: Get Word16
	profCapset <- get
	profTicks <- get
	profStackDepth <- get
	profCcsStack <- VU.replicateM (fromIntegral profStackDepth) get
	assert
	((fromIntegral payloadLen :: Int) == sum
	[ 4
	, 8 -- ticks
	, 1 -- stack depth
	, fromIntegral profStackDepth * 4
	])
	(return ())
	return $! ProfSampleCostCentre {..} ]

https://github.com/ghc/ghc/blob/73b22ff196160036ac10b762bf3a363fa8a451ad/rts/eventlog/EventLog.c#L1317-L1340 has

void postProfSampleCostCentre(Capability *cap,
                              CostCentreStack *stack,
                              StgWord64 tick)
{
    /* SNIP */
    postEventHeader(&eventBuf, EVENT_PROF_SAMPLE_COST_CENTRE);
    postPayloadSize(&eventBuf, len);
    postWord32(&eventBuf, cap->no);                                                /* THIS IS THE NOTABLE LINE */
    postWord64(&eventBuf, tick);
    /* SNIP */
}

• This seems to be an inconsistency between ghc master and ghc-events master -- I haven't done the due diligence to check wheter ghc master has changed this recently, eg.

• Or perhaps there's an implicit assumption that CapNo i is the sole member of Capset i?

So either there's a mismatch that should be fixed or I think it's deserving of a comment on profCapset. Thanks!

[maoe]

Yeah that indeed looks like a mismatch that should be fixed.

@mpickering Apparently you wrote both implementations in 9f00d38 and in ghc/ghc@17987a4. Was this intentional?

[mpickering]

I think it's probably a mistake.