On Thu 28-03-24 09:49:59, Kemeng Shi wrote: > on 3/27/2024 5:33 PM, Jan Kara wrote: > > On Thu 21-03-24 15:12:21, Kemeng Shi wrote: > >> > >> > >> on 3/20/2024 11:15 PM, Tejun Heo wrote: > >>> Hello, > >>> > >>> On Wed, Mar 20, 2024 at 07:02:22PM +0800, Kemeng Shi wrote: > >>>> We never use gdtc->dom set with GDTC_INIT_NO_WB, just remove unneeded > >>>> GDTC_INIT_NO_WB > >>>> > >>>> Signed-off-by: Kemeng Shi <shikemeng@xxxxxxxxxxxxxxx> > >>> ... > >>>> void global_dirty_limits(unsigned long *pbackground, unsigned long *pdirty) > >>>> { > >>>> - struct dirty_throttle_control gdtc = { GDTC_INIT_NO_WB }; > >>>> + struct dirty_throttle_control gdtc = { }; > >>> > >>> Even if it's currently not referenced, wouldn't it still be better to always > >>> guarantee that a dtc's dom is always initialized? I'm not sure what we get > >>> by removing this. > >> As we explicitly use GDTC_INIT_NO_WB to set global_wb_domain before > >> calculating dirty limit with domain_dirty_limits, I intuitively think the > >> dirty limit calculation in domain_dirty_limits is related to > >> global_wb_domain when CONFIG_WRITEBACK_CGROUP is enabled while the truth > >> is not. So this is a little confusing to me. > > > Hi Jan, > > I'm not sure I understand your confusion. domain_dirty_limits() calculates > > the dirty limit (and background dirty limit) for the dirty_throttle_control > > passed in. If you pass dtc initialized with GDTC_INIT[_NO_WB], it will > > compute global dirty limits. If the dtc passed in is initialized with > > MDTC_INIT() it will compute cgroup specific dirty limits. > No doubt about this. > > > > Now because domain_dirty_limits() does not scale the limits based on each > > device throughput - that is done only later in __wb_calc_thresh() to avoid> relatively expensive computations when we don't need them - and also > > because the effective dirty limit (dtc->dom->dirty_limit) is not updated by > > domain_dirty_limits(), domain_dirty_limits() does not need dtc->dom at all. > Acutally, here is the thing confusing me. For wb_calc_thresh, we always pass > dtc initialized with a wb (GDTC_INIT(wb) or MDTC_INIT(wb,..). The dtc > initialized with _NO_WB is only passed to domain_dirty_limits. However, The > dom initialized by _NO_WB for domain_dirty_limits is not needed at all. > > But that is a technical detail of implementation and I don't want this > > technical detail to be relied on by even more code. > Yes, I agree with this. So I wonder if it's acceptable to simply define > GDTC_INIT_NO_WB to empty for now instead of remove defination of > GDTC_INIT_NO_WB. When implementation of domain_dirty_limits() or any > other low level function in future using GDTC_INIT(_NO_WB) changes to > need dtc->domain, we re-define GDTC_INIT_NO_WB to proper value. > As this only looks confusing to me. I will drop this one in next version > if you still prefer to keep definatino of GDTC_INIT_NO_WB in the old way. Yeah, please keep the code as is for now. I agree this needs some cleanups but what you suggest is IMHO not an improvement. Honza -- Jan Kara <jack@xxxxxxxx> SUSE Labs, CR