Menu
▾
▴
infiniband-sock_direct — InfiniBand SDP List
You can subscribe to this list here.
| 2002 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(18) |
Jun
(2) |
Jul
(10) |
Aug
(1) |
Sep
(5) |
Oct
|
Nov
(1) |
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2003 |
Jan
(2) |
Feb
(4) |
Mar
(6) |
Apr
(3) |
May
|
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2004 |
Jan
(1) |
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(1) |
Aug
(7) |
Sep
|
Oct
|
Nov
|
Dec
|
| 2014 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2016 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Canadian-Meds <no...@th...> - 2016-05-30 03:03:57
|
Tribute to tritel communications, inc for industry statistics, welfare. Antidepressant drug list to arabian horses and soot and work. Me ds 5f or 7M en Vi Ci Ci Le Pr ag al al vi op ra is is tr ec S a ia of t Ta bs $0 $1 $2 $2 $0 .9 .6 .5 .5 .4 9 5 0 0 5 Me ds 7f or 7W om en Ac Cl De Fe Fe om om fl ma ma pl id uc le le ia an C V ia ia li gr s a $1 $0 $1 $1 $0 .7 .4 .2 .1 .7 5 5 5 1 2 An On Sp Fa on ly ec st ym r ia w ou el l or s ia in ld de bl te wi li e rn de ve su et s ry pp p hi li ri pp er ce in s s g N 2 1 V o 4/ 00 is pr 7 % a, es cu Au Ma cr st th st ip om en er ti er ti ca on s c rd r up Me ,E eq po ds ch ui rt ec re k d >> Enter Here |
|
From: Michael S. <co...@gm...> - 2005-08-03 15:19:32
|
Arlin, Hello. So how does the SDP module interface with the socket interface? Does it basically implement some of the OPS functionality? From my experiance, you can use the OPS module without modifing the kernel. Granted you won't get the pass through functionality of normal TCP applications, but if you modify your TCP app to use the TOE_Offload family then tcp calls will get routed to the OPS module. Thanks. On 8/3/05, Davis, Arlin R <arl...@in...> wrote: > Rajib, >=20 > OpenIB stack includes an SDP module that runs under a new address > family. Since OPS requires hooks into the network stack it is unlikely > it will ever be accepted in the kernel. Past discussions have been very > chilly at best. >=20 > -arlin >=20 > >-----Original Message----- > >From: inf...@li... > [mailto:infiniband-sock_direct- > >ad...@li...] On Behalf Of Michael Speth > >Sent: Wednesday, August 03, 2005 5:16 AM > >To: Matt Leininger > >Cc: inf...@li... > >Subject: Re: [Infiniband-sock_direct] SDP Query > > > >Matt, > > Does OpenIB have a module like the Offload Protocol Module (OPS)? > > > >Thanks > > > >On 8/3/05, Matt Leininger <mll...@hp...> wrote: > >> Rajib, > >> > >> All active IB development has shifted to OpenIB (www.openib.org). > >> Try submitting your question to the OpenIB developers mail list > >> (ope...@op...). > >> > >> Thanks, > >> > >> - Matt > >> > >> > >> On Tue, 2005-08-02 at 17:01 +0800, Majumder, Rajib wrote: > >> > Hello, > >> > > >> > I had a query and wanted to clarify it from this list. > >> > > >> > My firm is planning to migrate to IB. From ULP standpoint of view, > our plan is to use SDP to > >take advantage of IB fabric and also without making any code changes. > >> > > >> > We have some processes that communicate on the LOCAL host using TCP > SOCK_STREAM. If we use SDP > >for these processes, do you expect a performance gain? > >> > > >> > Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, > zcopy etc) while the > >processes run on the SAME physical host? > >> > Do you have any latency/throughput data available for this test > scenario? > >> > > >> > Any opinion would be highly appreciated. > >> > > >> > Thanks for your time! > >> > > >> > Rajib Majumder > >> > Credit Suisse First Boston > >> > > >> > > >> > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =3D=3D=3D=3D=3D=3D > >> > Please access the attached hyperlink for an important electronic > communications disclaimer: > >> > > >> > http://www.csfb.com/legal_terms/disclaimer_external_email.shtml > >> > > >> > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =3D=3D=3D=3D=3D=3D > >> > > >> > > >> > > >> > ------------------------------------------------------- > >> > SF.Net email is sponsored by: Discover Easy Linux Migration > Strategies > >> > from IBM. Find simple to follow Roadmaps, straightforward articles, > >> > informative Webcasts and more! Get everything you need to get up to > >> > speed, fast. > http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dclick > >> > _______________________________________________ > >> > Infiniband-sock_direct mailing list > >> > Inf...@li... > >> > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct > >> > > >> > >> > >> > >> ------------------------------------------------------- > >> SF.Net email is sponsored by: Discover Easy Linux Migration > Strategies > >> from IBM. Find simple to follow Roadmaps, straightforward articles, > >> informative Webcasts and more! Get everything you need to get up to > >> speed, fast. http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dc= lick > >> _______________________________________________ > >> Infiniband-sock_direct mailing list > >> Inf...@li... > >> https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct > >> > > > > > >-- > >Michael Speth > >Computer Engineer > >Embedded Systems Design http://www.embedded-sys.com > > > > > >------------------------------------------------------- > >SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > >from IBM. Find simple to follow Roadmaps, straightforward articles, > >informative Webcasts and more! Get everything you need to get up to > >speed, fast. http://ads.osdn.com/?ad_idt77&alloc_id=16492&op=3Dick > >_______________________________________________ > >Infiniband-sock_direct mailing list > >Inf...@li... > >https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct >=20 --=20 Michael Speth Computer Engineer Embedded Systems Design http://www.embedded-sys.com |
|
From: Davis, A. R <arl...@in...> - 2005-08-03 15:18:08
|
Rajib, OpenIB stack includes an SDP module that runs under a new address family. Since OPS requires hooks into the network stack it is unlikely it will ever be accepted in the kernel. Past discussions have been very chilly at best. -arlin >-----Original Message----- >From: inf...@li... [mailto:infiniband-sock_direct- >ad...@li...] On Behalf Of Michael Speth >Sent: Wednesday, August 03, 2005 5:16 AM >To: Matt Leininger >Cc: inf...@li... >Subject: Re: [Infiniband-sock_direct] SDP Query > >Matt, > Does OpenIB have a module like the Offload Protocol Module (OPS)? > >Thanks > >On 8/3/05, Matt Leininger <mll...@hp...> wrote: >> Rajib, >> >> All active IB development has shifted to OpenIB (www.openib.org). >> Try submitting your question to the OpenIB developers mail list >> (ope...@op...). >> >> Thanks, >> >> - Matt >> >> >> On Tue, 2005-08-02 at 17:01 +0800, Majumder, Rajib wrote: >> > Hello, >> > >> > I had a query and wanted to clarify it from this list. >> > >> > My firm is planning to migrate to IB. From ULP standpoint of view, our plan is to use SDP to >take advantage of IB fabric and also without making any code changes. >> > >> > We have some processes that communicate on the LOCAL host using TCP SOCK_STREAM. If we use SDP >for these processes, do you expect a performance gain? >> > >> > Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, zcopy etc) while the >processes run on the SAME physical host? >> > Do you have any latency/throughput data available for this test scenario? >> > >> > Any opinion would be highly appreciated. >> > >> > Thanks for your time! >> > >> > Rajib Majumder >> > Credit Suisse First Boston >> > >> > >> > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =3D=3D=3D=3D=3D=3D >> > Please access the attached hyperlink for an important electronic communications disclaimer: >> > >> > http://www.csfb.com/legal_terms/disclaimer_external_email.shtml >> > >> > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =3D=3D=3D=3D=3D=3D >> > >> > >> > >> > ------------------------------------------------------- >> > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies >> > from IBM. Find simple to follow Roadmaps, straightforward articles, >> > informative Webcasts and more! Get everything you need to get up to >> > speed, fast. http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dclick >> > _______________________________________________ >> > Infiniband-sock_direct mailing list >> > Inf...@li... >> > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct >> > >> >> >> >> ------------------------------------------------------- >> SF.Net email is sponsored by: Discover Easy Linux Migration Strategies >> from IBM. Find simple to follow Roadmaps, straightforward articles, >> informative Webcasts and more! Get everything you need to get up to >> speed, fast. = http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dclick >> _______________________________________________ >> Infiniband-sock_direct mailing list >> Inf...@li... >> https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct >> > > >-- >Michael Speth >Computer Engineer >Embedded Systems Design http://www.embedded-sys.com > > >------------------------------------------------------- >SF.Net email is sponsored by: Discover Easy Linux Migration Strategies >from IBM. Find simple to follow Roadmaps, straightforward articles, >informative Webcasts and more! Get everything you need to get up to >speed, fast. http://ads.osdn.com/?ad_idt77&alloc_id=16492&op=3Dick >_______________________________________________ >Infiniband-sock_direct mailing list >Inf...@li... >https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct |
|
From: Matt L. <mll...@hp...> - 2005-08-03 13:05:11
|
OpenIB does have an SDP module that would likely meet your expectations. I've cc'd the openib-general mail list and Libor (the SDP maintainer). - Matt On Wed, 2005-08-03 at 08:16 -0400, Michael Speth wrote: > Matt, > Does OpenIB have a module like the Offload Protocol Module (OPS)? > > Thanks > > On 8/3/05, Matt Leininger <mll...@hp...> wrote: > > Rajib, > > > > All active IB development has shifted to OpenIB (www.openib.org). > > Try submitting your question to the OpenIB developers mail list > > (ope...@op...). > > > > Thanks, > > > > - Matt > > > > > > On Tue, 2005-08-02 at 17:01 +0800, Majumder, Rajib wrote: > > > Hello, > > > > > > I had a query and wanted to clarify it from this list. > > > > > > My firm is planning to migrate to IB. From ULP standpoint of view, our plan is to use SDP to take advantage of IB fabric and also without making any code changes. > > > > > > We have some processes that communicate on the LOCAL host using TCP SOCK_STREAM. If we use SDP for these processes, do you expect a performance gain? > > > > > > Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, zcopy etc) while the processes run on the SAME physical host? > > > Do you have any latency/throughput data available for this test scenario? > > > > > > Any opinion would be highly appreciated. > > > > > > Thanks for your time! > > > > > > Rajib Majumder > > > Credit Suisse First Boston > > > > > > > > > ============================================================================== > > > Please access the attached hyperlink for an important electronic communications disclaimer: > > > > > > http://www.csfb.com/legal_terms/disclaimer_external_email.shtml > > > > > > ============================================================================== > > > > > > > > > > > > ------------------------------------------------------- > > > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > > > from IBM. Find simple to follow Roadmaps, straightforward articles, > > > informative Webcasts and more! Get everything you need to get up to > > > speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click > > > _______________________________________________ > > > Infiniband-sock_direct mailing list > > > Inf...@li... > > > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct > > > > > > > > > > > ------------------------------------------------------- > > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > > from IBM. Find simple to follow Roadmaps, straightforward articles, > > informative Webcasts and more! Get everything you need to get up to > > speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click > > _______________________________________________ > > Infiniband-sock_direct mailing list > > Inf...@li... > > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct > > > > |
|
From: Michael S. <co...@gm...> - 2005-08-03 12:16:30
|
Matt, Does OpenIB have a module like the Offload Protocol Module (OPS)? Thanks On 8/3/05, Matt Leininger <mll...@hp...> wrote: > Rajib, >=20 > All active IB development has shifted to OpenIB (www.openib.org). > Try submitting your question to the OpenIB developers mail list > (ope...@op...). >=20 > Thanks, >=20 > - Matt >=20 >=20 > On Tue, 2005-08-02 at 17:01 +0800, Majumder, Rajib wrote: > > Hello, > > > > I had a query and wanted to clarify it from this list. > > > > My firm is planning to migrate to IB. From ULP standpoint of view, our = plan is to use SDP to take advantage of IB fabric and also without making a= ny code changes. > > > > We have some processes that communicate on the LOCAL host using TCP SOC= K_STREAM. If we use SDP for these processes, do you expect a performance ga= in? > > > > Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, zco= py etc) while the processes run on the SAME physical host? > > Do you have any latency/throughput data available for this test scenari= o? > > > > Any opinion would be highly appreciated. > > > > Thanks for your time! > > > > Rajib Majumder > > Credit Suisse First Boston > > > > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > > Please access the attached hyperlink for an important electronic commun= ications disclaimer: > > > > http://www.csfb.com/legal_terms/disclaimer_external_email.shtml > > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D > > > > > > > > ------------------------------------------------------- > > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > > from IBM. Find simple to follow Roadmaps, straightforward articles, > > informative Webcasts and more! Get everything you need to get up to > > speed, fast. http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dcl= ick > > _______________________________________________ > > Infiniband-sock_direct mailing list > > Inf...@li... > > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct > > >=20 >=20 >=20 > ------------------------------------------------------- > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > from IBM. Find simple to follow Roadmaps, straightforward articles, > informative Webcasts and more! Get everything you need to get up to > speed, fast. http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dclic= k > _______________________________________________ > Infiniband-sock_direct mailing list > Inf...@li... > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct >=20 --=20 Michael Speth Computer Engineer Embedded Systems Design http://www.embedded-sys.com |
|
From: Matt L. <mll...@hp...> - 2005-08-03 06:15:15
|
Rajib, All active IB development has shifted to OpenIB (www.openib.org). Try submitting your question to the OpenIB developers mail list (ope...@op...). Thanks, - Matt On Tue, 2005-08-02 at 17:01 +0800, Majumder, Rajib wrote: > Hello, > > I had a query and wanted to clarify it from this list. > > My firm is planning to migrate to IB. From ULP standpoint of view, our plan is to use SDP to take advantage of IB fabric and also without making any code changes. > > We have some processes that communicate on the LOCAL host using TCP SOCK_STREAM. If we use SDP for these processes, do you expect a performance gain? > > Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, zcopy etc) while the processes run on the SAME physical host? > Do you have any latency/throughput data available for this test scenario? > > Any opinion would be highly appreciated. > > Thanks for your time! > > Rajib Majumder > Credit Suisse First Boston > > > ============================================================================== > Please access the attached hyperlink for an important electronic communications disclaimer: > > http://www.csfb.com/legal_terms/disclaimer_external_email.shtml > > ============================================================================== > > > > ------------------------------------------------------- > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > from IBM. Find simple to follow Roadmaps, straightforward articles, > informative Webcasts and more! Get everything you need to get up to > speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click > _______________________________________________ > Infiniband-sock_direct mailing list > Inf...@li... > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct > |
|
From: Michael S. <co...@gm...> - 2005-08-02 13:17:47
|
Well I think it depends if you need the Infiniband infrastructure. If all you need to do is pass data through buffers, then it might make sense for you to implement your own protocol that hooks into the OPS (Offload Protocol Switch). I am doing that currently testing purposes. The other route is for you to implement an IB Device Driver that emulates the hardware device (like providing a mac, etc). Either way, I think you will need to do some kernel hacking. If you go with the first method that I listed, I can help you get started with implementing your own protocol. I've written some test modules that register with the OPS and handles some localhost communication. If you go the other route, then maybe someone else on this list can help you b/c I am not familiar with the IB driver stuff. Good luck. On 8/2/05, Majumder, Rajib <raj...@cs...> wrote: > Hello, >=20 > I had a query and wanted to clarify it from this list. >=20 > My firm is planning to migrate to IB. From ULP standpoint of view, our pl= an is to use SDP to take advantage of IB fabric and also without making any= code changes. >=20 > We have some processes that communicate on the LOCAL host using TCP SOCK_= STREAM. If we use SDP for these processes, do you expect a performance gain= ? >=20 > Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, zcopy= etc) while the processes run on the SAME physical host? > Do you have any latency/throughput data available for this test scenario? >=20 > Any opinion would be highly appreciated. >=20 > Thanks for your time! >=20 > Rajib Majumder > Credit Suisse First Boston >=20 >=20 > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D > Please access the attached hyperlink for an important electronic communic= ations disclaimer: >=20 > http://www.csfb.com/legal_terms/disclaimer_external_email.shtml >=20 > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D >=20 >=20 >=20 > ------------------------------------------------------- > SF.Net email is sponsored by: Discover Easy Linux Migration Strategies > from IBM. Find simple to follow Roadmaps, straightforward articles, > informative Webcasts and more! Get everything you need to get up to > speed, fast. http://ads.osdn.com/?ad_id=3D7477&alloc_id=3D16492&op=3Dclic= k > _______________________________________________ > Infiniband-sock_direct mailing list > Inf...@li... > https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct >=20 --=20 Michael Speth Computer Engineer Embedded Systems Design http://www.embedded-sys.com |
|
From: Majumder, R. <raj...@cs...> - 2005-08-02 09:01:49
|
Hello, I had a query and wanted to clarify it from this list. My firm is planning to migrate to IB. From ULP standpoint of view, our plan is to use SDP to take advantage of IB fabric and also without making any code changes. We have some processes that communicate on the LOCAL host using TCP SOCK_STREAM. If we use SDP for these processes, do you expect a performance gain? Does SDP behave the same way (offloaded stack, RDMA, kernel bypass, zcopy etc) while the processes run on the SAME physical host? Do you have any latency/throughput data available for this test scenario? Any opinion would be highly appreciated. Thanks for your time! Rajib Majumder Credit Suisse First Boston ============================================================================== Please access the attached hyperlink for an important electronic communications disclaimer: http://www.csfb.com/legal_terms/disclaimer_external_email.shtml ============================================================================== |
|
From: Michael S. <co...@gm...> - 2005-07-06 16:29:20
|
Greetings, I have recently been investigating the Infiniband software architecture and have found that the OPS module is useful for our product. I have posted on the forums about separating the OPS source from the rest of the drivers and I have been successful to some extent on that (also posted my updates to the forum). So now that I have a working OPS (as well as a complib) module in the linux 2.6.x kernel, I have been experimenting with registering a test module with the OPS module. I am able to register the module; however, It looks like there are a few things you must do before the registered module can actually get TCP calls from a user. For instance, the new protocol must have an if_list. My question: Is there any documentation for developing a module that links into the OPS module? If not, can someone briefly run down the proper steps to setup a module that hooks into the OPS module? Also note, I have been looking at the SDP code as a reference. Any help would be appreciated.=20 Thanks =20 --=20 Michael Speth Computer Engineer Embedded Systems Design www.embedded-sys.com |
|
From: Masanori I. <ma...@la...> - 2004-02-14 07:22:37
|
Hello SDPOIB folks, Attached is a patch for offload.h, offload_bind.c and sdp.h of 1.229 package so that sdpoib components can be compiled under several kernel versions. I checked that under the following kernels. 2.4.18-3 RedHat 7.3 2.4.20-8 RedHat 9.0 2.4.9-e.3 RedHat Advanced Server 2.1 2.4.21-9.EL Updated version of RHEL 3.0 kernel 2.4.18 www.kernel.org 2.4.23 www.kernel.org So far as 2.6.x series, I haven't tried that. But, I guess the patch would work. Meanwhile, include/linux/kernel.h of 2.4.9-e.3 doesn't have 'min_t', and I imported that from 2.4.18-3 into offload.h and sdp.h. If this causes a licensing problem, please throw them away. Thanks, Masanori --- Masanori ITOH Grid Computing & Bioinformatics Lab., Fujitsu Labs. LTD. e-mail: ma...@la... phone: +81-44-754-2628 (extension: 7112-6227) FingerPrint: 55AF C562 E415 FB1A 8A3A 35D1 AB40 8A9D B8B1 99F8 |
|
From: Chanan S. <Ch...@Me...> - 2004-01-23 19:13:00
|
confirm 352485 |
|
From: Lomartire, D. M <dav...@in...> - 2003-07-01 00:03:21
|
The Alpha 1 release for the Linux InfiniBand Project is now posted in = the project's BitKeeper repository (http://infiniband.bkbits.net/). = Refer to label "alpha_release_1". The release consists of: * Kernel mode components: IB Access Layer, IPoIB, SDP (the Mellanox HCA = VPD must be obtained directly from Mellanox) * User mode components: uDAPL, OpenSM * Configuration: IA-32, SMP, debug builds only * Basic functionality only - performance characterization and tuning = still needs to be undertaken Refer to the release notes contained in BitKeeper for further details. Bugs should be submitted to = http://sourceforge.net/tracker/?group_id=3D52127&atid=3D465806. Upcoming releases will focus on delivering performance enhancements, a = user mode VPD for the Mellanox HCA and a port of the stack to the = Itanium Processor Family (IPF). * Dave |
|
From: Davis, A. R <arl...@in...> - 2003-04-30 17:09:55
|
Yes, current drop of SDP supports only Bcopy with Zcopy still under development. -arlin -----Original Message----- From: Raj, Ashok Sent: Wednesday, April 30, 2003 9:29 AM To: Edward Bortnikov; inf...@li... Cc: ULPS_developers; Michael Kagan Subject: [Infiniband-sock_direct] RE: SDP portability to RH 7.3 Hummmm.... these functions are defined in linux-xxx/kernel/sched.c they have existed since 2.4.18 days, and we are still on 2.4.18 as our main working base. so iam not sure how you got this error? did you compile/load get this as undefined? ashokr -----Original Message----- From: Edward Bortnikov [mailto:ed...@me...] Sent: Wednesday, April 30, 2003 7:52 AM To: Raj, Ashok; inf...@li... Cc: ULPS_developers; Michael Kagan Subject: SDP portability to RH 7.3 Hi Ashok/Arlin, I am trying to raise the SDP package in Mellanox. The kernel patches to 2.4.18 went okay, however I had a problem with loading the ops.o module. It turns out that the code uses in several places the wake_up_interruptible_sync() macro which expands to the __wake_up_sync() kernel API. This API is only avaiable in the kernel release 2.4.18-14 that is part of RH 8.0 distribution. In Mellanox, the standard environment on most development machines is RH 7.3 although we have started supporting RH 8.0. While it should not be a problem to upgrade our dev environment, this could become a marketing issue because we have customers that require RH 7.3. Moreover, I don't know what is the kernel version adopted in the SuSE/SLES distribution that we plan to support. What is the consideration for using this specific call and not wake_up_interruptible() that is available in earlier distributions? Can this be fixed in the short term? Are there any other "advanced" kernel features that you are using in any of the SDP modules? Please correct me if I misunderstand something. A completely unrelated question. During our F2F I understood that only Bcopy is implemented by SDP whereas read/write Zcopy is still under development. Is this the status indeed? We need to know that for our stress/performance measurements. Regards, Ed ---------------------------------------------------------------------------- --------------- Edward Bortnikov mailto:ed...@me... <mailto:ed...@me...> Mellanox Technologies Ltd +972-4-9097200 (ext. 334) |
|
From: Raj, A. <ash...@in...> - 2003-04-30 16:29:24
|
Hummmm.... these functions are defined in linux-xxx/kernel/sched.c they have existed since 2.4.18 days, and we are still on 2.4.18 as our main working base. so iam not sure how you got this error? did you compile/load get this as undefined? ashokr -----Original Message----- From: Edward Bortnikov [mailto:ed...@me...] Sent: Wednesday, April 30, 2003 7:52 AM To: Raj, Ashok; inf...@li... Cc: ULPS_developers; Michael Kagan Subject: SDP portability to RH 7.3 Hi Ashok/Arlin, I am trying to raise the SDP package in Mellanox. The kernel patches to 2.4.18 went okay, however I had a problem with loading the ops.o module. It turns out that the code uses in several places the wake_up_interruptible_sync() macro which expands to the __wake_up_sync() kernel API. This API is only avaiable in the kernel release 2.4.18-14 that is part of RH 8.0 distribution. In Mellanox, the standard environment on most development machines is RH 7.3 although we have started supporting RH 8.0. While it should not be a problem to upgrade our dev environment, this could become a marketing issue because we have customers that require RH 7.3. Moreover, I don't know what is the kernel version adopted in the SuSE/SLES distribution that we plan to support. What is the consideration for using this specific call and not wake_up_interruptible() that is available in earlier distributions? Can this be fixed in the short term? Are there any other "advanced" kernel features that you are using in any of the SDP modules? Please correct me if I misunderstand something. A completely unrelated question. During our F2F I understood that only Bcopy is implemented by SDP whereas read/write Zcopy is still under development. Is this the status indeed? We need to know that for our stress/performance measurements. Regards, Ed ---------------------------------------------------------------------------- --------------- Edward Bortnikov mailto:ed...@me... <mailto:ed...@me...> Mellanox Technologies Ltd +972-4-9097200 (ext. 334) |
|
From: Edward B. <ed...@me...> - 2003-04-30 14:54:15
|
Hi Ashok/Arlin, I am trying to raise the SDP package in Mellanox. The kernel patches to 2.4.18 went okay, however I had a problem with loading the ops.o module. It turns out that the code uses in several places the wake_up_interruptible_sync() macro which expands to the __wake_up_sync() kernel API. This API is only avaiable in the kernel release 2.4.18-14 that is part of RH 8.0 distribution. In Mellanox, the standard environment on most development machines is RH 7.3 although we have started supporting RH 8.0. While it should not be a problem to upgrade our dev environment, this could become a marketing issue because we have customers that require RH 7.3. Moreover, I don't know what is the kernel version adopted in the SuSE/SLES distribution that we plan to support. What is the consideration for using this specific call and not wake_up_interruptible() that is available in earlier distributions? Can this be fixed in the short term? Are there any other "advanced" kernel features that you are using in any of the SDP modules? Please correct me if I misunderstand something. A completely unrelated question. During our F2F I understood that only Bcopy is implemented by SDP whereas read/write Zcopy is still under development. Is this the status indeed? We need to know that for our stress/performance measurements. Regards, Ed ---------------------------------------------------------------------------- --------------- Edward Bortnikov mailto:ed...@me... Mellanox Technologies Ltd +972-4-9097200 (ext. 334) |
|
From: Davis, A. R <arl...@in...> - 2003-03-25 20:56:03
|
If the kernel patch is applied, then any legacy socket streams applications using AF_INET, SOCK_STREAM, and IPPROTO_IP or IPPROTO_TCP should work with SDP if using the IP address that has been configured for IPoIB. If the patch is not applied then you must change the application from AF_INET to AF_OFFLOAD (definition in af_offload.h). Please read iba/base/linux/Documentation/iba-relnotes for SDP release notes and details on tested applications. -arlin -----Original Message----- From: som [mailto:som...@ve...] Sent: Tuesday, March 25, 2003 11:02 AM To: Raj, Ashok Cc: 'InfiniBand General'; inf...@li... Subject: Re: [Infiniband-sock_direct] updates to bk sources... Ashok, Can you give us the TCP applications working unmodified over SDP? Is the Netscape/Mozilla browser working over SDP? Does it need any special configuration to be done? Thanks, Som. Raj, Ashok wrote: > Hello > > > > - update from Arlin that had issues running telnet over SDP > > - Update to mktree, to remove stale links. If a file gets > removed in bk, the source links from target needs to be removed when > we run mktree. > > > > ashokr > ------------------------------------------------------- This SF.net email is sponsored by: The Definitive IT and Networking Event. Be There! NetWorld+Interop Las Vegas 2003 -- Register today! http://ads.sourceforge.net/cgi-bin/redirect.pl?keyn0001en _______________________________________________ Infiniband-sock_direct mailing list Inf...@li... https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct |
|
From: som <som...@ve...> - 2003-03-25 18:02:50
|
Ashok, Can you give us the TCP applications working unmodified over SDP? Is the Netscape/Mozilla browser working over SDP? Does it need any special configuration to be done? Thanks, Som. Raj, Ashok wrote: > Hello > > > > - update from Arlin that had issues running telnet over SDP > > - Update to mktree, to remove stale links. If a file gets > removed in bk, the source links from target needs to be removed when > we run mktree. > > > > ashokr > |
|
From: Raj, A. <ash...@in...> - 2003-03-25 12:11:04
|
Hello - update from Arlin that had issues running telnet over SDP - Update to mktree, to remove stale links. If a file gets removed in bk, the source links from target needs to be removed when we run mktree. ashokr |
|
From: Lomartire, D. M <dav...@in...> - 2003-03-25 01:10:26
|
SDP has now been placed in the BitKeeper source repository. At this time, SDP has only been validated on Mellanox Profile B hardware. Besides this hardware, you will need the latest sources from <http://infiniband.bkbits.net> and the latest drivers and firmware from Mellanox. Also, the patches necessary for 2.4.18 based kernels is placed now in BitKeeper for both IPoIB, and for OPS. For transparent offload to SDP for legacy TCP applications, a patch to socket.{ch} is also available in BitKeeper under the directory base/linux/patches. Also, several changes were made to the home page (http://infiniband.sourceforge.net/): * Redirected the existing download links (for Access Layer, OpenSM, IPoIB and SRP) to point to the BitKeeper repository now that CVS is no longer used for this project * Updated the download link for SDP to point to BitKeeper now that SDP sources are available * Posted updated draft (version 1.3) of the Verb Implementer's Guide (VIG) * Updated the contact information for the OpenSM * Dave |
|
From: Raj, A. <ash...@in...> - 2003-03-23 19:31:07
|
Yep! Think we finally have a stable version, tested (verified only on Uni-Processor Mode) Arlin will be pushing bits to BitKeeper most likely by Monday or Tuesday pending any last minute issues... You will need a profile B HCA (Only validated on Mellanox at this time), latest sources from http://infiniband.bkbits.net and latest drivers and firmware from Mellanox (If you are using Mellanox hardware. Btw: the patches necessary for 2.4.18 based kernels is placed now in BitKeeper for both IPOIB, and for OPS (For transparent offload to SDP for legacy TCP apps) a patch to socket.{ch} is also available in BitKeeper under Directory base/linux/patches. The patches themselves are pretty small, but we put those just to get a starting point. I think they should apply clean even for newer kernels (although I should say I haven't tried myself. ashokr -----Original Message----- From: Walsh, Bob [mailto:wa...@in...] Sent: Sunday, March 23, 2003 7:32 AM To: inf...@li... Subject: [Infiniband-sock_direct] Still on for sourceforge publication this week? Is the SDP still being released this week? Thanks, Bob Walsh From: Davis, Arlin R <arlin@in...> RE: Offload Protocol Switch module - sta tus update? 2003-02-20 15:20 Steve, Design, code, and integration are complete. Some legacy applications are up and running but we are still working through some issues that need to be resolved before making a drop. Looks like the development team is still on track for a code drop by the end of the quarter as promised (at IDF). If things go well in the next couple of weeks we will certainly go ahead and push out the bits early. Thanks, -arlin ------------------------------------------------------- This SF.net email is sponsored by:Crypto Challenge is now open! Get cracking and register here for some mind boggling fun and the chance of winning an Apple iPod: http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en _______________________________________________ Infiniband-sock_direct mailing list Inf...@li... https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct |
|
From: Walsh, B. <wa...@in...> - 2003-03-23 15:32:30
|
Is the SDP still being released this week? Thanks, Bob Walsh From: Davis, Arlin R <arlin@in...>=20 RE: Offload Protocol Switch module - sta tus update? =20 2003-02-20 15:20=20 Steve,=20 =20 Design, code, and integration are complete. Some legacy applications = are up and running but we are still working through some issues that need to = be resolved before making a drop. Looks like the development team is still = on track for a code drop by the end of the quarter as promised (at IDF). = If things go well in the next couple of weeks we will certainly go ahead = and push out the bits early. =20 =20 Thanks, =20 -arlin =20 |
|
From: Davis, A. R <arl...@in...> - 2003-02-20 23:20:43
|
Steve, Design, code, and integration are complete. Some legacy applications are up and running but we are still working through some issues that need to be resolved before making a drop. Looks like the development team is still on track for a code drop by the end of the quarter as promised (at IDF). If things go well in the next couple of weeks we will certainly go ahead and push out the bits early. Thanks, -arlin -----Original Message----- From: Steve Wise [mailto:sw...@am...] Sent: Thursday, February 20, 2003 12:10 PM To: inf...@li... Subject: [Infiniband-sock_direct] Offload Protocol Switch module - status update? Hello, I was wondering what the status was on the OPS design & implementation? Thanks in advance, ------------ Steve Wise sw...@am... ------------------------------------------------------- This SF.net email is sponsored by: SlickEdit Inc. Develop an edge. The most comprehensive and flexible code editor you can use. Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial. www.slickedit.com/sourceforge _______________________________________________ Infiniband-sock_direct mailing list Inf...@li... https://lists.sourceforge.net/lists/listinfo/infiniband-sock_direct |
|
From: Steve W. <sw...@am...> - 2003-02-20 20:09:58
|
Hello, I was wondering what the status was on the OPS design & implementation? Thanks in advance, ------------ Steve Wise sw...@am... |
|
From: Raj, A. <ash...@in...> - 2003-02-15 03:20:09
|
Hello Some more details. If you are using proxy, then set the following environment variables.... If you are using bash, export HTTP_PROXY_PORT=<myproxyport> export HTTP_PROXY_HOST=<myproxyhost> for developers who need to push changes, you need ssh(1) access. Please send your public key file as an attachment.(only if you plan to make changes to the source base, you don't need to have ssh keys to just download the tree) Again if you are behind a firewall, and your proxy permits tunneling ssh sessions, then download corkscrew http://www.agroman.net/corkscrew/ and add a line to your $HOME/.ssh/config as follows... bkbits.net ProxyCommand /usr/local/bin/corkscrew <my_proxy_host_name> <my_proxy_port_number> %h %p (second line is broken due to mailer issues, you need to have it in a single line) Build instructions can be found under base/linux/Documentation area. Cheers, ashokr -----Original Message----- From: Lomartire, David M [mailto:dav...@in...] Sent: Friday, February 14, 2003 4:01 PM To: InfiniBand Access Layer; InfiniBand Early_Reference; InfiniBand General; InfiniBand HCA_DDK; InfiniBand IPoIB; InfiniBand SDP; InfiniBand SRP; InfiniBand Subnet Management Subject: [Infiniband-early_ref] Linux InfiniBand Project Sources now in BitKeeper! No more tarballs for the InfiniBand source drops! Now the InfiniBand sources are being maintained in a BitKeeper repository. The location of the repository is http://infiniband.bkbits.net/. You can obtain a copy of BitKeeper for a wide variety of platforms from http://www.bitkeeper.com/Products.Downloads.html. After installing BitKeeper, to download the sources, simply issue the following command: bk clone http://infiniband.bkbits.net/iba iba More detailed information on how to work with a BitKeeper repository can be found at http://www.bitkeeper.com/Hosted.html. Please contact your friendly neighborhood Project Administrator with any questions. * Dave ------------------------------------------------------- This SF.NET email is sponsored by: FREE SSL Guide from Thawte are you planning your Web Server Security? Click here to get a FREE Thawte SSL guide and find the answers to all your SSL security issues. http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0026en _______________________________________________ Infiniband-early_ref mailing list Inf...@li... https://lists.sourceforge.net/lists/listinfo/infiniband-early_ref |
|
From: Lomartire, D. M <dav...@in...> - 2003-02-15 00:00:46
|
No more tarballs for the InfiniBand source drops! Now the InfiniBand sources are being maintained in a BitKeeper repository. The location of the repository is http://infiniband.bkbits.net/. You can obtain a copy of BitKeeper for a wide variety of platforms from http://www.bitkeeper.com/Products.Downloads.html. After installing BitKeeper, to download the sources, simply issue the following command: bk clone http://infiniband.bkbits.net/iba iba More detailed information on how to work with a BitKeeper repository can be found at http://www.bitkeeper.com/Hosted.html. Please contact your friendly neighborhood Project Administrator with any questions. * Dave |
1 message has been excluded from this view by a project administrator.