Re: [PATCH] scsi: core: Fix missing lock when read async_scan in Scsi_Host

Next message: Mark Brown: "Re: [PATCH net-next 2/2] net: sfp: manage receiver and transmitter regulators"
Previous message: Markus Schneider-Pargmann (TI): "[PATCH 4/6] arm64: dts: ti: k3-am62p5-sk: Split r5f memory region"
In reply to: Damien Le Moal: "Re: [PATCH] scsi: core: Fix missing lock when read async_scan in Scsi_Host"
Next in thread: James Bottomley: "Re: [PATCH] scsi: core: Fix missing lock when read async_scan in Scsi_Host"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

From: Christoph Hellwig

Date: Tue Mar 03 2026 - 10:19:50 EST

No, using a lock to read a single scalar value makes zero sense.
Just switch to READ_ONCE/WRITE_ONCE to make alpha and KCSAN happy.