[PATCH v2 0/5] Venus fatal error handling · Stanimir Varbanov <hidden> · 2021-06-08
[PATCH v2 2/5] venus: Make sys_error flag an atomic bitops · Stanimir Varbanov <hidden> · 2021-06-08
RE: [PATCH v2 2/5] venus: Make sys_error flag an atomic bitops · David Laight <hidden> · 2021-06-09
[PATCH v2 3/5] venus: hfi: Check for sys error on session hfi functions · Stanimir Varbanov <hidden> · 2021-06-08
[PATCH v2 5/5] venus: Handle fatal errors during encoding and decoding · Stanimir Varbanov <hidden> · 2021-06-08
[PATCH v2 1/5] venus: venc: Use pmruntime autosuspend · Stanimir Varbanov <hidden> · 2021-06-08
[PATCH v2 4/5] venus: helpers: Add helper to mark fatal vb2 error · Stanimir Varbanov <hidden> · 2021-06-08
Re: [PATCH v2 0/5] Venus fatal error handling · <hidden> · 2021-10-06

RE: [PATCH v2 2/5] venus: Make sys_error flag an atomic bitops

From: David Laight <hidden>
Date: 2021-06-09 16:03:13
Also in: linux-arm-msm, lkml

From: Stanimir Varbanov <redacted>

Sent: 08 June 2021 12:42

Make the sys_error flag an atomic bitops in order to avoid
locking in sys_error readers.

I don't think that makes any difference at all.

It is only atomic w.r.t other bitops in the same bitmap.
Even if it contained the fullest memory barrier you can
imagine the result of test_bit() is immediately stale.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help