From mboxrd@z Thu Jan 1 00:00:00 1970 From: erik quanstrom Date: Tue, 6 Mar 2012 13:55:49 -0500 To: 9p-st@imu.li, 9fans@9fans.net Message-ID: <48dfd79dc45c0d5e9d88c6a79de7202e@brasstown.quanstro.net> In-Reply-To: <2579b9.228186c7.D0Vx.mx@tumtum.plumbweb.net> References: <2579b6.8126cb52.62GS.mx@tumtum.plumbweb.net> <504d1e9e6a08271d15827ea4c00507ff@chula.quanstro.net> <2579b9.032a3487.pkFh.mx@tumtum.plumbweb.net> <2579b9.228186c7.D0Vx.mx@tumtum.plumbweb.net> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Subject: Re: [9fans] usb slowness Topicbox-Message-UUID: 66d2210c-ead7-11e9-9d60-3106f5b1d025 On Tue Mar 6 10:37:06 EST 2012, 9p-st@imu.li wrote: > > my working theory is that i'm getting all the interrupts, but not soon > > enough... the fascinating piece is that the olpc and the pc with intel > > ehc take just about the same amount of time. > > indeed, if i set the interrupt threshold control to 1 micro-frame (they > default to 8), i get 10-11 sec dd, say, half as slow. and ping time to my > switch is half. and i can get about twice the throughput from the file > server, which is still kinda slow. > > (hget yields about 6Mbit over a 100Mbit link) it sounds like you have it all figured out. but just in case there is a spurious interrupt causing trouble, you can check with this rather embarassing acid script. sorry, i can't get acid to recoginze intrcountx as a vlong. acid -l acid -k 1 /386/9pccpu acid: i=0; while i<256*2 do {if(intrcountx[i]>0) then {print(i/2\D, "\t", intrcountx[i]);} i=i+2;} 7 143 14 44355 50 99090099 65 19938 73 37571 145 9814 153 1 161 42 169 7 - erik