Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-09-10 Thread Michael Ellerman
Michael Ellerman writes: > On Tue, 28 Jul 2020 12:37:41 -0500, Nathan Lynch wrote: >> The drmem lmb list can have hundreds of thousands of entries, and >> unfortunately lookups take the form of linear searches. As long as >> this is the case, traversals have the potential to monopolize the CPU >>

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-09-09 Thread Michael Ellerman
On Tue, 28 Jul 2020 12:37:41 -0500, Nathan Lynch wrote: > The drmem lmb list can have hundreds of thousands of entries, and > unfortunately lookups take the form of linear searches. As long as > this is the case, traversals have the potential to monopolize the CPU > and provoke lockup reports, work

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-08-11 Thread Nathan Lynch
Nathan Lynch writes: > Michael Ellerman writes: >> One thought, which I possibly should not put in writing, is that we >> could use the alignment of the pointer as a poor man's substitute for a >> counter, eg: >> >> +static inline struct drmem_lmb *drmem_lmb_next(struct drmem_lmb *lmb) >> +{ >> +

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-08-10 Thread Nathan Lynch
Michael Ellerman writes: > One thought, which I possibly should not put in writing, is that we > could use the alignment of the pointer as a poor man's substitute for a > counter, eg: > > +static inline struct drmem_lmb *drmem_lmb_next(struct drmem_lmb *lmb) > +{ > + if (lmb % PAGE_SIZE == 0)

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-08-02 Thread Michael Ellerman
Nathan Lynch writes: > Michael Ellerman writes: >> Nathan Lynch writes: >>> Michael Ellerman writes: Nathan Lynch writes: > Laurent Dufour writes: >> Le 28/07/2020 à 19:37, Nathan Lynch a écrit : >>> The drmem lmb list can have hundreds of thousands of entries, and >>> un

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-31 Thread Nathan Lynch
Michael Ellerman writes: > Nathan Lynch writes: >> Michael Ellerman writes: >>> Nathan Lynch writes: Laurent Dufour writes: > Le 28/07/2020 à 19:37, Nathan Lynch a écrit : >> The drmem lmb list can have hundreds of thousands of entries, and >> unfortunately lookups take the f

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-31 Thread Michael Ellerman
Nathan Lynch writes: > Michael Ellerman writes: >> Nathan Lynch writes: >>> Laurent Dufour writes: Le 28/07/2020 à 19:37, Nathan Lynch a écrit : > The drmem lmb list can have hundreds of thousands of entries, and > unfortunately lookups take the form of linear searches. As long as

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-30 Thread Nathan Lynch
Michael Ellerman writes: > Nathan Lynch writes: >> Laurent Dufour writes: >>> Le 28/07/2020 à 19:37, Nathan Lynch a écrit : The drmem lmb list can have hundreds of thousands of entries, and unfortunately lookups take the form of linear searches. As long as this is the case, traver

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-29 Thread Michael Ellerman
Nathan Lynch writes: > Laurent Dufour writes: >> Le 28/07/2020 à 19:37, Nathan Lynch a écrit : >>> The drmem lmb list can have hundreds of thousands of entries, and >>> unfortunately lookups take the form of linear searches. As long as >>> this is the case, traversals have the potential to monopo

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-28 Thread Nathan Lynch
Hi Laurent, Laurent Dufour writes: > Le 28/07/2020 à 19:37, Nathan Lynch a écrit : >> The drmem lmb list can have hundreds of thousands of entries, and >> unfortunately lookups take the form of linear searches. As long as >> this is the case, traversals have the potential to monopolize the CPU >>

Re: [PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-28 Thread Laurent Dufour
Le 28/07/2020 à 19:37, Nathan Lynch a écrit : The drmem lmb list can have hundreds of thousands of entries, and unfortunately lookups take the form of linear searches. As long as this is the case, traversals have the potential to monopolize the CPU and provoke lockup reports, workqueue stalls, an

[PATCH] powerpc/pseries: explicitly reschedule during drmem_lmb list traversal

2020-07-28 Thread Nathan Lynch
The drmem lmb list can have hundreds of thousands of entries, and unfortunately lookups take the form of linear searches. As long as this is the case, traversals have the potential to monopolize the CPU and provoke lockup reports, workqueue stalls, and the like unless they explicitly yield. Rather