cfq-iosched: remove dead_key from cfq_io_context
in [Kernel]
|
Prev: perf, trace: Use per-tracepoint-per-cpu hlist to track events
Next: perf, trace: Use per-tracepoint-per-cpu hlist to track events
From: Jens Axboe on 21 May 2010 05:50 On Thu, May 20 2010, Konstantin Khlebnikov wrote: > Remove ->dead_key field from cfq_io_context to shrink its size to 128 bytes. > (64 bytes for 32-bit hosts) > > Use lower bit in ->key as dead-mark, instead of moving key to separate field. > After this for dead cfq_io_context we got cic->key != cfqd automatically. > Thus, io_context's last-hit cache should work without changing. > > Now to check ->key for non-dead state compare it with cfqd, > instead of checking ->key for non-null value as it was before. > > Plus remove obsolete race protection in cfq_cic_lookup. > This race gone after v2.6.24-1728-g4ac845a This, and the second patch, look really good. How much testing have you done with it? -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Konstantin Khlebnikov on 21 May 2010 11:00 On 21.05.2010 13:40, Jens Axboe wrote: > On Thu, May 20 2010, Konstantin Khlebnikov wrote: >> Remove ->dead_key field from cfq_io_context to shrink its size to 128 bytes. >> (64 bytes for 32-bit hosts) >> >> Use lower bit in ->key as dead-mark, instead of moving key to separate field. >> After this for dead cfq_io_context we got cic->key != cfqd automatically. >> Thus, io_context's last-hit cache should work without changing. >> >> Now to check ->key for non-dead state compare it with cfqd, >> instead of checking ->key for non-null value as it was before. >> >> Plus remove obsolete race protection in cfq_cic_lookup. >> This race gone after v2.6.24-1728-g4ac845a > > This, and the second patch, look really good. How much testing have you > done with it? > I have run multiple fsstress and simultaneously switch io-scheduler in loop for several hours, no oopses and leaks detected. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Jens Axboe on 21 May 2010 13:40
On Fri, May 21 2010, Konstantin Khlebnikov wrote: > On 21.05.2010 13:40, Jens Axboe wrote: >> On Thu, May 20 2010, Konstantin Khlebnikov wrote: >>> Remove ->dead_key field from cfq_io_context to shrink its size to 128 bytes. >>> (64 bytes for 32-bit hosts) >>> >>> Use lower bit in ->key as dead-mark, instead of moving key to separate field. >>> After this for dead cfq_io_context we got cic->key != cfqd automatically. >>> Thus, io_context's last-hit cache should work without changing. >>> >>> Now to check ->key for non-dead state compare it with cfqd, >>> instead of checking ->key for non-null value as it was before. >>> >>> Plus remove obsolete race protection in cfq_cic_lookup. >>> This race gone after v2.6.24-1728-g4ac845a >> >> This, and the second patch, look really good. How much testing have you >> done with it? >> > > I have run multiple fsstress and simultaneously switch io-scheduler in loop > for several hours, no oopses and leaks detected. OK, I'll queue it up for later in the 2.6.35 cycle. -- Jens Axboe -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |