On (08/02/18 14:13), Andrew Morton wrote:
[..]
> That changelog is rather hard to follow. Please review my edits:
>
> : If zram supports writeback feature, it's no longer a BD_CAP_SYNCHRONOUS_IO
^BDI_CAP_SYNCHRONOUS_IO
[..]
> A reader looking at this would wonder "why the heck are we doing that".
> Adding a code comment would help them.
The interesting thing here is that include/linux/backing-dev.h
BDI_CAP_SYNCHRONOUS_IO comment says
"Device is so fast that asynchronous IO would be inefficient."
Which is not the reason why BDI_CAP_SYNCHRONOUS_IO is used by ZRAM.
Probably, the comment needs to be updated as well.
Both SWP_SYNCHRONOUS_IO and BDI_CAP_SYNCHRONOUS_IO tend to pivot
"efficiency" [looking at the comments], but in ZRAM's case the whole
reason to use SYNC IO is a race condition and user-after-free that
follows.
-ss
