Re: -CURRENT slowdown in last 2 weeks
On Sat, Feb 24, 2001 at 10:50:17AM +0200, Mark Murray wrote: > > and now performance is very good, event with: > > > > kern.random.sys.harvest_ethernet: 1 > > kern.random.sys.harvest_point_to_point: 0 > > kern.random.sys.harvest_interrupt: 1 > > You mean "even with"? If so, then I am very pleased indeed! Yeah, that's what I mean! Granted, this is just my workstation so I don't have that many interrupts, but I often have a lot of network traffic, and ATA drives generate quite a lot of interrupts during make world ;-) Bye, Andrea -- Yes, I've heard of "decaf." What's your point? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: -CURRENT slowdown in last 2 weeks
> and now performance is very good, event with: > > kern.random.sys.harvest_ethernet: 1 > kern.random.sys.harvest_point_to_point: 0 > kern.random.sys.harvest_interrupt: 1 You mean "even with"? If so, then I am very pleased indeed! M -- Mark Murray Warning: this .sig is umop ap!sdn To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
[Solved] Re: -CURRENT slowdown in last 2 weeks
> > Do you have MUTEX_DEBUG in your kernel? > Sorry guys, my bad. As John and Kris reminded me, the slowdown was because of this - should have checked again the archive, I remember it had been mentioned before. Bye, Andrea -- It is easier to fix Unix than to live with NT. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: -CURRENT slowdown in last 2 weeks
On Tue, Feb 20, 2001 at 01:50:23AM +0100, Andrea Campi wrote: > I am noticing a severe slowdown on my -CURRENT system. It actually started > after Feb [3-5] changes in intrupt handling, but I didn't really notice > until I run a make world (which I delayed doing because of, well you guess, > the libc breakage). When I say severe I mean make buildworld takes x3 longer. Do you have MUTEX_DEBUG in your kernel? Kris PGP signature
Re: -CURRENT slowdown in last 2 weeks
> > Hmm, Feb 3-5 (looks). > > You mean the preemptive scheduling committed on Feb 1? Can you try updating > to early this week to see if it goes away? Hi John, every "recent" version I tried resulted in a slowdown so I didn't recompile very often; until tonight, I was still runninng a kernel from 20010214. Tonight I created a new kernel but I finally figured out I should take out most debugging options: options DDB # options WITNESS # options WITNESS_DDB # options MUTEX_DEBUG # options INVARIANT_SUPPORT # options INVARIANTS and now performance is very good, event with: kern.random.sys.harvest_ethernet: 1 kern.random.sys.harvest_point_to_point: 0 kern.random.sys.harvest_interrupt: 1 Later tonight (my locale) I saw a few other commits to ithreads and such, exp. your commit which should fix my issue with pccard (thanks a lot for that!), so later a will update again and start building kernels adding back the debugging options one at a time. I'll post the result. Bye, Andrea -- Actually, Microsoft is sort of a mixture between the Borg and the Ferengi. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: -CURRENT slowdown in last 2 weeks
On Wed, Feb 21, 2001 at 10:20:46PM +, Pierre Y. Dampure wrote: > Andrea Campi wrote: > > > I am noticing a severe slowdown on my -CURRENT system. It actually started > > after Feb [3-5] changes in intrupt handling, but I didn't really notice > > until I run a make world (which I delayed doing because of, well you guess, > > the libc breakage). When I say severe I mean make buildworld takes x3 longer. > > Hmmm. Buit world yesterday evening on a 733 VAIO, UDMA ATA drive, took around > 1h15mn, which is fairly much what I would expect... U using SCSI? Nope, UDMA33 (IBM-DARA-20600 on IBM Thinkpad). Question: you built a world yesterday, but what world did you have BEFORE? I mean, what matters is the kernel/world which was running while you were compiling, was that recent (< 15 days old)? Are you seeing any lock reversal message? Thanks anyway, bye, Andrea -- Reboot America. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: -CURRENT slowdown in last 2 weeks
Andrea Campi wrote: > Nope, UDMA33 (IBM-DARA-20600 on IBM Thinkpad). > > Question: you built a world yesterday, but what world did you have BEFORE? I > mean, what matters is the kernel/world which was running while you were > compiling, was that recent (< 15 days old)? Are you seeing any lock reversal > message? > The previous world and kernel where from that same day (I had been playing around with user.h and kmod.mk and rebuilt to make sure everything was still okay). I did not see any lock reversal messages. My config is pretty standard, let me know if you want a peek (the only "non-standard" bit is the fact that I'm running newcard, I doubt this would matter in this case). Best Regards, PYD To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
RE: -CURRENT slowdown in last 2 weeks
On 20-Feb-01 Andrea Campi wrote: > I am noticing a severe slowdown on my -CURRENT system. It actually started > after Feb [3-5] changes in intrupt handling, but I didn't really notice > until I run a make world (which I delayed doing because of, well you guess, > the libc breakage). When I say severe I mean make buildworld takes x3 longer. Hmm, Feb 3-5 (looks). You mean the preemptive scheduling committed on Feb 1? Can you try updating to early this week to see if it goes away? -- John Baldwin <[EMAIL PROTECTED]> -- http://www.FreeBSD.org/~jhb/ PGP Key: http://www.baldwin.cx/~john/pgpkey.asc "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: -CURRENT slowdown in last 2 weeks
Andrea Campi wrote: > I am noticing a severe slowdown on my -CURRENT system. It actually started > after Feb [3-5] changes in intrupt handling, but I didn't really notice > until I run a make world (which I delayed doing because of, well you guess, > the libc breakage). When I say severe I mean make buildworld takes x3 longer. Hmmm. Buit world yesterday evening on a 733 VAIO, UDMA ATA drive, took around 1h15mn, which is fairly much what I would expect... U using SCSI? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message