From mboxrd@z Thu Jan  1 00:00:00 1970
From: Tristan <9p-st@imu.li>
To: 9fans@9fans.net
Content-Type: text/plain; charset=UTF-8
Message-Id: <2579b9.87e726c6.np4k.mx@tumtum.plumbweb.net>
In-Reply-To: <2579b9.4d6b93e5.B75p.mx@tumtum.plumbweb.net>
References: <a8a4d75adaf927ca6d650b46859510c5@hamnavoe.com><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>
	<48dfd79dc45c0d5e9d88c6a79de7202e@brasstown.quanstro.net>
	<2579b9.4d6b93e5.B75p.mx@tumtum.plumbweb.net>
Date: Wed,  7 Mar 2012 00:50:37 -0500
User-Agent: mx-alpha
Content-Transfer-Encoding: quoted-printable
Subject: Re: [9fans] usb slowness
Topicbox-Message-UUID: 67697b60-ead7-11e9-9d60-3106f5b1d025

a little reading later.

the usb mass storage bulk only spec stipulates that only one request may
be active at a time.

wince.


so, bigger buffer? yup, 32k reads double the speed again. 64k reads
aren't quite right, something clips it's tail. now i'm up to ~11 MB/s.

except that it crashes the pc and olpc kernels. kw does fine though (for
now!). changing devmnt seem unlikely to be very safe without checking
on every fileserver in case it now happily assumes 8k (plus headers) is
enough for anything.

anybody messed with devmnt's Msgsize successfully?

enjoy,
tristan

--=20
All original matter is hereby placed immediately under the public domain.