[openfabrics-ewg] Current OFED kernel snapshot - problems in back porting SRP to RH4
Ishai Rabinovitz
ishai at mellanox.co.il
Tue May 2 01:30:13 PDT 2006
On Tue, May 02, 2006 at 10:34:29AM +0300, Aviram Gutman wrote:
>
> Yes, we have uploaded a pre version of RC4. You can find it in:
>
> URL: https://openib.org/svn/gen2/branches/1.0/ofed/releases
>
> OFED-1.0-rc4-pre4.tgz
> OFED-1.0-rc4-pre4.tgz.md5sum
>
> Please pay attention that we still face issues. The following is the diff between RC3 and the pre:
>
> 1. Bug fixes according to problems reported.
>
> 2. SRP - with new features: FMR, tunable parameters, SRP daemon - We have an issue with the back port of SRP to RH4 U2 and U3. Ishai will issue a mail with explanation.
>
> 3. Open MPI - new package based on 1.1a3 - Please be noted that RPM building process failed. Vlad will
>
> 4. RDS - new version from main trunk
>
> 5. Kernel code based on git
>
> 6. Standard network configuration
>
>
> Known issues:
> 1. ipath installation fails on 2.6.9 - 2.6.11* kernels
> 2. OSU MPI compilation fails on SLES10, PPC64
> 3. SRP is not supported on 2.6.9 - 2.6.13* kernels - Ishai will follow up with details
> 4. Open MPI RPM build process fails - Jeff, will you be able to send us fixes by Wed?
>
>
> Regards,
> Aviram
>
> -----Original Message-----
> From: openfabrics-ewg-bounces at openib.org [mailto:openfabrics-ewg-bounces at openib.org] On Behalf Of Doug Ledford
> Sent: Monday, May 01, 2006 11:53 PM
> To: openfabrics-ewg at openib.org
> Subject: [openfabrics-ewg] Current OFED kernel snapshot
>
> There's rumored to be a significant number of changes between rc3 and rc4 on the kernel module front. I would like to get started on integrating those changes sooner rather than later. So, where would I go to get a snapshot of the latest OFED kernel
> tree. So far I've only found kernel trees under the tags directory and obviously the rc4 tag hasn't been populated yet.
>
> --
> Doug Ledford <dledford at redhat.com>
> Red Hat, Inc.
> 1801 Varsity Dr.
> Raleigh, NC 27606
>
> _______________________________________________
> openfabrics-ewg mailing list
> openfabrics-ewg at openib.org
> http://openib.org/mailman/listinfo/openfabrics-ewg
Hi
We have a problem when trying to back port SRP to RH4 U2 and U3 (Actually to
any kernel earlier than 2.6.13).
The problem is when the SCSI driver is calling to eh_abort_handler,
or to eh_device_reset_handler.
In the current kernel (starting from 2.6.13) this call is made without
host_lock spin-lock locked.
In the SRP code that performs the abort and the reset (srp_send_tsk_mgmt) we
send a message to the target and we wait for a response from the target.
In early versions of the kernel the SCSI driver performs irq_spinlock_save to
the host_lock before calling to the abort or reset handlers.
This creates a problem: The SRP driver can not go to sleep until the target will
answer.
Any ideas?
--
Ishai Rabinovitz
More information about the ewg
mailing list