Messages in this thread | | | Date | Sat, 19 Jun 2004 13:09:58 +0200 | From | Francois Romieu <> | Subject | Re: wmb versus smp_wmb |
| |
Margit Schubert-While <margitsw@t-online.de> : > As the $SUBJECT implies, when should one use > wmb() versus smp_wmb() ?
If the code which must see the variable(s) modified before the wmb() can run on a separate CPU, then it should be a smp_wmb().
For instance host H1 wants to update A then B and host H2 needs to be sure that if it reads the updated value of B, then it reads the updated value of A as well. B could be an event that uses a different channel instead of a memory update.
Usually (for me :o) ), the issue between the two (or more) CPUs is complicated by the fact that there is some device behind a PCI bus whose behavior depends on the same data as well.
Suggested reading: Schimmel + Aspirin recommended use.
-- Ueimor - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |