c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2026-05-23 10:54:35 +03:00
Code Issues Packages Projects Releases Wiki Activity

DDT: Reduce global DDT lock scope during writes

Browse Source 
Before this change DDT lock was taken 4 times per written block,
and as effectively a pool-wide lock it can be highly congested.
This change introduces a new per-entry dde_io_lock, protecting some
fields during I/O ready and done stages, so that we don't need the
global lock there.

According to my write tests on 64-thread system with 4KB blocks this
significantly reduce the global lock contention, reducing CPU usage
from 100% to expected ~80%, and increasing write throughput by 10%.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Rob Norris <robn@despairlabs.com>
Signed-off-by: Alexander Motin <alexander.motin@TrueNAS.com>
Closes #17960
This commit is contained in:
Alexander Motin
2025-12-01 13:44:10 -05:00
committed by GitHub
parent a5b665df39
commit 928eccc5bc
3 changed files with 87 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files
module/zfs/ddt.c
+2
View File
@@ -1010,6 +1010,7 @@ ddt_alloc_entry_io(ddt_entry_t *dde)
return;
dde->dde_io = kmem_zalloc(sizeof (ddt_entry_io_t), KM_SLEEP);
mutex_init(&dde->dde_io->dde_io_lock, NULL, MUTEX_DEFAULT, NULL);
}
static void
@@ -1022,6 +1023,7 @@ ddt_free(const ddt_t *ddt, ddt_entry_t *dde)
if (dde->dde_io->dde_repair_abd != NULL)
abd_free(dde->dde_io->dde_repair_abd);
mutex_destroy(&dde->dde_io->dde_io_lock);
kmem_free(dde->dde_io, sizeof (ddt_entry_io_t));
}