Lockless patterns: full memory barriers

Posted Mar 12, 2021 17:03 UTC (Fri) by firolwn (guest, #96711)
Parent article: Lockless patterns: full memory barriers

Hi Paolo, great article. I think maybe you forgot to add 'smp_mb();' to the two diagrams which are just around the line starting with 'Due to transitivity'.
--
Firo

Lockless patterns: full memory barriers

Posted Mar 15, 2021 14:35 UTC (Mon) by firolwn (guest, #96711) [Link] (1 responses)

After reading 'Multicopy atomicity' section from kernel Documentation/memory-barriers.txt, I realize that I am wrong and no more smp_mb() is necessary to add.
--
Firo

Lockless patterns: full memory barriers

Posted Apr 16, 2021 7:01 UTC (Fri) by pbonzini (subscriber, #60935) [Link]

It's not intuitive at all, but only one memory barrier matters in each of the two cases. But both are needed (separately) to ensure that x=0 && y = 0 is not possible.