On Thu, 21 Apr 2022 02:13:39 -0700, Dylan Yudaken wrote:
> This series addresses a rare but real error condition when a CQE is
> dropped. Many applications rely on 1 SQE resulting in 1 CQE and may even
> block waiting for the CQE. In overflow conditions if the GFP_ATOMIC
> allocation fails, the CQE is dropped and a counter is incremented. However
> the application is not actively signalled that something bad has
> happened. We would like to indicate this error condition to the
> application but in a way that does not rely on the application doing
> invasive changes such as checking a flag before each wait.
>
> [...]
Applied, thanks!
[1/6] io_uring: add trace support for CQE overflow
commit: f457ab8deb017140aef05be3027a00a18a7d16b7
[2/6] io_uring: trace cqe overflows
commit: 2a847e6faf76810ae68a6e81bd9ac3a7c81534d0
[3/6] io_uring: rework io_uring_enter to simplify return value
commit: db9bb58b391c9e62da68bc139598e8470d892c77
[4/6] io_uring: use constants for cq_overflow bitfield
commit: b293240e2634b2100196d7314aeeb84299ce6d5b
[5/6] io_uring: return an error when cqe is dropped
commit: 34a7ee8a42c8496632465f3f0b444b3a7b908c46
[6/6] io_uring: allow NOP opcode in IOPOLL mode
commit: ebbe59f49556822b9bcc7b0d4d96bae31f522905
Best regards,
--
Jens Axboe
