Skip to contentSkip to conversation

Question on CE_DESC_FLAGS_META_DATA field.

DISCUSSION Ben Greear 2 participants 3 messages Jan 9, 2015 – Jan 12, 2015 Qualcomm Atheros ATH10K wireless driver

lore
Share & tools
lore

Ben Greear

author
#0 lore raw

While poking in the CE code, I notice this below.

Shouldn't the CE_DESC_FLAGS_META_DATA_LSB be 2, and if not, can someone explain why?

[greearb@ben-dt2 ath10k]$ grep CE_DESC_FLAGS_META_DATA *h ce.h:#define CE_DESC_FLAGS_META_DATA_MASK 0xFFFC ce.h:#define CE_DESC_FLAGS_META_DATA_LSB 3

and this from core.h:

#define MS(_v, _f) (((_v) & _f##_MASK) >> _f##_LSB)

Thanks,
Ben

Signature · 8 lines
Ben Greear <[email protected]>
Candela Technologies Inc  http://www.candelatech.com


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k
#1 lore raw
Quoted 8 lines · Ben Greear

While poking in the CE code, I notice this below.

Shouldn't the CE_DESC_FLAGS_META_DATA_LSB be 2, and if not, can someone explain why?

[greearb@ben-dt2 ath10k]$ grep CE_DESC_FLAGS_META_DATA *h ce.h:#define CE_DESC_FLAGS_META_DATA_MASK 0xFFFC ce.h:#define CE_DESC_FLAGS_META_DATA_LSB 3

Nice catch!

This shouldn't have been harmful though. Meta data part is used to store htc endpoint ids which shouldn't overflow the mask even with lsb=3 and as far as I know neither CE nor firmware cares what is stored there.

Michał

_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

author
#2 lore raw
Quoted 10 lines · Michal Kazior
Quoted 8 lines · Ben Greear

While poking in the CE code, I notice this below.

Shouldn't the CE_DESC_FLAGS_META_DATA_LSB be 2, and if not, can someone explain why?

[greearb@ben-dt2 ath10k]$ grep CE_DESC_FLAGS_META_DATA *h ce.h:#define CE_DESC_FLAGS_META_DATA_MASK 0xFFFC ce.h:#define CE_DESC_FLAGS_META_DATA_LSB 3

Quoted 6 lines

Nice catch!

This shouldn't have been harmful though. Meta data part is used to store htc endpoint ids which shouldn't overflow the mask even with lsb=3 and as far as I know neither CE nor firmware cares what is stored there.

Ok, I'll fix this in my tree and post a patch sometime soon.

Upstream 10.1 firmware does ignore the transfer_id, but I am fixing that in my firmware to aid debugging (and fixing ath10k driver to use it for debugging as well).

Thanks,
Ben

Signature · 8 lines
Ben Greear <[email protected]>
Candela Technologies Inc  http://www.candelatech.com


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k