Message ID | 20250304144346.1025658-1-arnd@kernel.org (mailing list archive) |
---|---|
State | New |
Headers | show |
Series | firmware: arm_scmi: use ioread64() instead of ioread64_hi_lo() | expand |
Hi Arnd, On Tue, Mar 04, 2025 at 03:43:34PM +0100, Arnd Bergmann wrote: > From: Arnd Bergmann <arnd@arndb.de> > > The scmi_common_fastchannel_db_ring() function calls either ioread64() > or ioread64_hi_lo() depending on whether it is compiler for 32-bit > or 64-bit architectures. > > The same logic is used to define ioread64() itself in the > linux/io-64-nonatomic-hi-lo.h header file, so the special case > is not really needed. > > The behavior here should not change at all. LGTM. I just sent PR for v6.15 this morning. Either I can redo PR with this included or you can just apply this directly. I am fine either way. If you go for the latter approach, Reviewed-by: Sudeep Holla <sudeep.holla@arm.com> If you prefer me to send updated PR, let me know. I don't want to create confusion by recalling by PR and sending an updated one without consulting you. -- Regards, Sudeep
diff --git a/drivers/firmware/arm_scmi/driver.c b/drivers/firmware/arm_scmi/driver.c index 60050da54bf2..1c75a4c9c371 100644 --- a/drivers/firmware/arm_scmi/driver.c +++ b/drivers/firmware/arm_scmi/driver.c @@ -1997,17 +1997,7 @@ static void scmi_common_fastchannel_db_ring(struct scmi_fc_db_info *db) else if (db->width == 4) SCMI_PROTO_FC_RING_DB(32); else /* db->width == 8 */ -#ifdef CONFIG_64BIT SCMI_PROTO_FC_RING_DB(64); -#else - { - u64 val = 0; - - if (db->mask) - val = ioread64_hi_lo(db->addr) & db->mask; - iowrite64_hi_lo(db->set | val, db->addr); - } -#endif } /**