On 7/29/21 9:10 AM, Bart Van Assche wrote:
On 7/29/21 1:03 AM, Bean Huo wrote:
How did you compare the performance gain/loss after reverting this
commit?
I measured the performance impact of patches 11 and 12 combined. In a 4
KB read IOPS test I see that these two patches combined improve
performance by 1%. This illustrates that the two MMIO accesses of the
UTRLCNR register are slower than the single MMIO access of the doorbell
register.
A correction: I wanted to refer to the combination of patches 11 and 13
instead of 11 and 12.
Bart.
