Date: Sat, 4 May 2013 14:18:20 -0700 From: Jack Vogel <jfvogel@gmail.com> To: Richard Sharpe <realrichardsharpe@gmail.com> Cc: FreeBSD Net <freebsd-net@freebsd.org>, Adrian Chadd <adrian@freebsd.org> Subject: Re: Is there any way to limit the amount of data in an mbuf chain submitted to a driver? Message-ID: <CAFOYbckUzavh5W7BoChg7Dg7cGPK7JLJ5O3Mc8syAYa9LQKcxg@mail.gmail.com> In-Reply-To: <CACyXjPy=ErKS_A3Nmwa8PnMc_D=2LByd4VJPFdTj9sqgSaCTfw@mail.gmail.com> References: <CACyXjPwC5LRb7DT82n6PMbawceER3_nHko9c9tvrdQqceLiPww@mail.gmail.com> <CAJ-Vmon_5eyXMP5UOsVVBP8UgKQLw5HLMO1NgswoGb-zF=2wtg@mail.gmail.com> <CACyXjPzu3fXpo0i5YcdVBFye%2BRFTPUye=fgZ%2BycTkkiEmcRh%2BQ@mail.gmail.com> <CAFOYbcmZMW=-7Mwz9mwJLeM3Ju%2BF8_AsXFAPqCa8%2BuuRWq3xsg@mail.gmail.com> <CACyXjPy=ErKS_A3Nmwa8PnMc_D=2LByd4VJPFdTj9sqgSaCTfw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Ahh, Twinville, new hardware :) The version at the tip is 2.5.8 and I am working on version 2.5.12 internally that I hope to commit next week... so your version is "a bit old" :) I would do some testing on newer code. Jack On Sat, May 4, 2013 at 1:54 PM, Richard Sharpe <realrichardsharpe@gmail.com>wrote: > On Sat, May 4, 2013 at 1:41 PM, Jack Vogel <jfvogel@gmail.com> wrote: > > If you don't use TSO you will hurt your TX performance significantly fr= om > > the tests that I've run. What exactly is the device you are using, I > don't > > have the source in front of me now, but I'm almost sure that the limit = is > > not 64K but 256K, or are you using some ancient version of the driver? > > ix0 pnpinfo vendor=3D0x8086 device=3D0x1528 subvendor=3D0x808= 6 > subdevice=3D0x0001 class=3D0x020000 at slot=3D0 function=3D0 > ix1 pnpinfo vendor=3D0x8086 device=3D0x1528 subvendor=3D0x808= 6 > subdevice=3D0x0001 class=3D0x020000 at slot=3D0 function=3D1 > > The version calls itself ixgbe-2.4.4 ... > > Hmmm, copyright is 2001-2010 ... so perhaps a bit old. > > > Jack > > > > > > > > On Sat, May 4, 2013 at 1:30 PM, Richard Sharpe < > realrichardsharpe@gmail.com> > > wrote: > >> > >> On Sat, May 4, 2013 at 10:39 AM, Adrian Chadd <adrian@freebsd.org> > wrote: > >> > On 4 May 2013 06:52, Richard Sharpe <realrichardsharpe@gmail.com> > wrote: > >> >> Hi folks, > >> >> > >> >> I understand better why I am seeing EINVAL intermittently when > sending > >> >> data from Samba via SMB2. > >> >> > >> >> The ixgbe driver, for TSO reasons, limits the amount of data that c= an > >> >> be DMA'd to 65535 bytes. It returns EINVAL for any mbuf chain large= r > >> >> than that. > >> >> > >> >> The SO_SNDBUF for that socket is set to 131972. Mostly there is les= s > >> >> than 64kiB of space available, so that is all TCP etc can put into > the > >> >> socket in one chain of mbufs. However, every now and then there is > >> >> more than 65535 bytes available in the socket buffers, and we have = an > >> >> SMB packet that is larger than 65535 bytes, and we get hit. > >> >> > >> >> To confirm this I am going to set SO_SNDBUF back to the default of > >> >> 65536 and test again. My repros are very reliable. > >> >> > >> >> However, I wondered if my only way around this if I want to continu= e > >> >> to use SO_SNDBUF sizes larger than 65536 is to fragment large mbuf > >> >> chains in the driver? > >> > > >> > Hm, is this is a problem without TSO? > >> > >> We are using the card without TSO, so I am thinking of changing that > >> limit to 131072 and retesting. > >> > >> I am currently testing with SO_SNDBUF=3D32768 and have not hit the > problem. > >> > >> > Is the problem that the NIC can't handle a frame that big, or a buff= er > >> > that big? > >> > Ie - if you handed the hardware two descriptors of 64k each, for the > >> > same IP datagram, will it complain? > >> > >> I can't find any documentation, but it seems that with TSO it cannot > >> handle a frame that big. Actually, since we are not using TSO, there > >> really should not be a problem with larger frames. > >> > >> > Or do you need to break it up into two separate IP datagrams, facing > >> > the driver, with a maximum size of 64k each? > >> > >> Not sure, but it looks like we need to do that. > >> > >> > >> -- > >> Regards, > >> Richard Sharpe > >> (=E4=BD=95=E4=BB=A5=E8=A7=A3=E6=86=82=EF=BC=9F=E5=94=AF=E6=9C=89=E6=9D= =9C=E5=BA=B7=E3=80=82--=E6=9B=B9=E6=93=8D) > >> _______________________________________________ > >> freebsd-net@freebsd.org mailing list > >> http://lists.freebsd.org/mailman/listinfo/freebsd-net > >> To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org" > > > > > > > > -- > Regards, > Richard Sharpe > (=E4=BD=95=E4=BB=A5=E8=A7=A3=E6=86=82=EF=BC=9F=E5=94=AF=E6=9C=89=E6=9D=9C= =E5=BA=B7=E3=80=82--=E6=9B=B9=E6=93=8D) >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAFOYbckUzavh5W7BoChg7Dg7cGPK7JLJ5O3Mc8syAYa9LQKcxg>