Hi Andrew, On Fri, Jun 13, 2025 at 11:04:15AM -0700, Andrew Morton wrote: > On Fri, 13 Jun 2025 23:26:33 +0900 Robert Pang <robertpang@xxxxxxxxxx> wrote: > > > Hi Andrew > > > > Bcache is designed to boost the I/O performance of slower storage > > (HDDs, network-attached storage) by leveraging fast SSDs as a block > > cache. This functionality is critical in significantly reducing I/O > > latency. Therefore, any notable increase in bcache's latency severely > > diminishes its value. For instance, our tests show a P100 (max) > > latency spike from 600 ms to 2.4 seconds every 5 minutes due to this > > regression. In real-world environments, this increase will cause > > frequent timeouts and stalls in end-user applications that rely on > > bcache's latency improvements, highlighting the urgent need to address > > this issue. > > Great, thanks. Let's please incorporate this into the v2 changelogging. > > > > > Also, if we are to address this regression in -stable kernels then > > > > reverting 866898efbb25 is an obvious way - it is far far safer. So > > > > please also tell us why the proposed patchset is a better way for us to > > > > go. > > > > > > > I agree that reverting 866898efbb25 is a much safer and smaller change > > > for backporting. In fact, I previously raised the discussion of whether > > > we should revert the commit or instead introduce an equality-aware API > > > and use it. The bcache maintainer preferred the latter, and I also > > > believe that it is a more forward-looking approach. Given that bcache > > > has run into this issue, it's likely that other users with similar use > > > cases may encounter it as well. We wouldn't want those users to > > > continue relying on the current default heapify behavior. So, although > > > reverting may be more suitable for stable in isolation, adding an > > > equality-aware API could better serve a broader set of use cases going > > > forward. > > "much safer and smaller" is very desirable for backporting, please. > After all, 866898efbb25 didn't really fix anything and reverting that > takes us back to a known-to-work implementation. > > I of course have no problem making the changes in this patchset for > "going forward"! > > So if agreeable, please prepare a patch which reverts 866898efbb25. > Robert's words above are a great basis for that patch's description. > Sure, I'll prepare a revert patch to address the issue and plan to submit it for backporting to -stable. However, I'd like to confirm whether the following patch series structure would be appropriate: - Patch 1: Revert 866898efbb25 and CC it to stable - Patch 2–8: Introduce the new equality-aware heap API - Patch 9: Revert Patch 1 and switch bcache to use the new API In this case, we would only backport Patch 1 to stable. Alternatively, would you prefer we simply revert 866898efbb25 without introducing and using the new API in the same series? Regards, Kuan-Wei
