Re: [PATCH v13 0/2] MR3504: ntdll: An implementation of SRWLOCK that closer matches Windows'.

[Yuxuan Shui (＠yshui)]

On Fri Aug 25 10:04:56 2023 +0000, Yuxuan Shui wrote:
> So I noticed another problem. I said I was worried about the releasing
> side (TO) being blocked, but my approach isn't better. If TO is in
> contention with many acquirers, it would have to keep chasing the tail
> of the list to update its head pointer, which can be growing
> indefinitely. This would behave almost the same as having a spinlock on
> the list.

@zfigura i think this is still salvageable, i have a rough idea now, i will try to flesh it out and update the patch. (unless i found holes before i finish it)