linux-rdma.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Laurence Oberman <loberman-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
To: Bart Van Assche <Bart.VanAssche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: Testing latest linux-next 4.9 ib_srp and ib_srpt
Date: Tue, 20 Dec 2016 16:09:37 -0500 (EST)	[thread overview]
Message-ID: <653622322.8201942.1482268177727.JavaMail.zimbra@redhat.com> (raw)
In-Reply-To: <1668746735.8200653.1482266682585.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>



----- Original Message -----
> From: "Laurence Oberman" <loberman-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> To: "Bart Van Assche" <Bart.VanAssche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
> Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Sent: Tuesday, December 20, 2016 3:44:42 PM
> Subject: Re: Testing latest linux-next 4.9 ib_srp and ib_srpt
> 
> 
> 
> ----- Original Message -----
> > From: "Laurence Oberman" <loberman-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> > To: "Bart Van Assche" <Bart.VanAssche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
> > Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > Sent: Tuesday, December 20, 2016 2:43:48 PM
> > Subject: Re: Testing latest linux-next 4.9 ib_srp and ib_srpt
> > 
> > 
> > 
> > ----- Original Message -----
> > > From: "Laurence Oberman" <loberman-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> > > To: "Bart Van Assche" <Bart.VanAssche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
> > > Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > > Sent: Tuesday, December 20, 2016 2:33:26 PM
> > > Subject: Re: Testing latest linux-next 4.9 ib_srp and ib_srpt
> > > 
> > > Hello Bart
> > > 
> > > I pulled the latest linux-next and built kernels for both server and
> > > client
> > > to rerun all my EDR tests for srp.
> > > 
> > > For some reason the I/O size is being capped again to 1MB in my testing.
> > > Using my same testbed.
> > > Remember we spent a lot of time making sure we could do 4MB I/O :)
> > > 
> > > Its working fine in the RHEL 7.3 kernel so before I start going back
> > > testing
> > > upstream kernels decided to ask.
> > > 
> > > Have you tested large I/O with latest linux-next
> > > 
> > > Server Configuration
> > > ---------------------
> > > Linux fedstorage.bos.redhat.com 4.9.0+
> > > 
> > > [root@fedstorage modprobe.d]# cat ib_srp.conf
> > > options ib_srp cmd_sg_entries=64 indirect_sg_entries=2048
> > > 
> > > [root@fedstorage modprobe.d]# cat ib_srpt.conf
> > > options ib_srpt srp_max_req_size=8296
> > > 
> > > Also Using
> > > 
> > > # Set the srp_sq_size
> > > for i in
> > > /sys/kernel/config/target/srpt/0xfe800000000000007cfe900300726e4e
> > > /sys/kernel/config/target/srpt/0xfe800000000000007cfe900300726e4f
> > > do
> > > 	echo 16384 > $i/tpgt_1/attrib/srp_sq_size
> > > done
> > > 
> > > Client Configuration
> > > --------------------
> > > Linux ibclient 4.9.0+
> > > 
> > > [root@ibclient modprobe.d]# cat ib_srp.conf
> > > options ib_srp cmd_sg_entries=255 indirect_sg_entries=2048
> > > 
> > > dd if=/dev/sdw bs=4096k of=/dev/null iflag=direct
> > > 
> > > ### RECORD    4 >>> ibclient <<< (1482261733.001) (Tue Dec 20 14:22:13
> > > 2016)
> > > ###
> > > # DISK STATISTICS (/sec)
> > > #
> > > <---------reads---------------><---------writes--------------><--------averages-------->
> > > Pct
> > > #Time     Name       KBytes Merged  IOs Size  Wait  KBytes Merged  IOs
> > > Size
> > > Wait  RWSize  QLen  Wait SvcTim Util
> > > 14:22:13 sdw         1373184      0 1341 1024     2       0      0    0
> > > 0
> > > 0    1024     3     2      0   97
> > > 
> > > 
> > > If I reboot into my 7.3 kernel its back to what I expect
> > > 
> > >  dd if=/dev/sdw bs=4096k of=/dev/null iflag=direct
> > > 
> > > 
> > > ### RECORD    3 >>> ibclient <<< (1482262254.001) (Tue Dec 20 14:30:54
> > > 2016)
> > > ###
> > > # DISK STATISTICS (/sec)
> > > #
> > > <---------reads---------------><---------writes--------------><--------averages-------->
> > > Pct
> > > #Time     Name       KBytes Merged  IOs Size  Wait  KBytes Merged  IOs
> > > Size
> > > Wait  RWSize  QLen  Wait SvcTim Util
> > > 14:30:54 sdw         172032    129   42 4096     3       0      0    0
> > > 0
> > > 0    4096     1     3      3   130
> > > 
> > > --
> > > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> > > the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > > 
> > 
> > Hi Bart,
> > 
> > Just FYI
> > 
> > That dd snap was just as I had stopped the dd.
> > 
> > Here is a stable dd snap with the RHEL kernel, I just noticed the merging,
> > need to reboot back into upstream to compare again.
> > No merging seen in upstream.
> > 
> > Thanks
> > Laurence
> > 
> > ### RECORD    6 >>> ibclient <<< (1482262723.001) (Tue Dec 20 14:38:43
> > 2016)
> > ###
> > # DISK STATISTICS (/sec)
> > #
> > <---------reads---------------><---------writes--------------><--------averages-------->
> > Pct
> > #Time     Name       KBytes Merged  IOs Size  Wait  KBytes Merged  IOs Size
> > Wait  RWSize  QLen  Wait SvcTim Util
> > 14:38:43 sdw         1200128    879  293 4096     3       0      0    0
> > 0
> > 0    4096     1     3      3   95
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> > the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > 
> 
> Replying to my own message to keep the thread going
> 
> 
> This is Linux ibclient 4.8.0-rc4
> 
> Behaves like I expected, and I see the 4MB I/O sizes. as I had already tested
> this.
> 
> dd if=/dev/sdw bs=4096k of=/dev/null iflag=direct
> 
> 
> ### RECORD    6 >>> ibclient <<< (1482266543.001) (Tue Dec 20 15:42:23 2016)
> ###
> # DISK STATISTICS (/sec)
> #
> <---------reads---------------><---------writes--------------><--------averages-------->
> Pct
> #Time     Name       KBytes Merged  IOs Size  Wait  KBytes Merged  IOs Size
> Wait  RWSize  QLen  Wait SvcTim Util
> 15:42:23 sdw         278528    201   68 4096     2       0      0    0    0
> 0    4096     1     2      2   206
> 
> Rebooting back into 4.9 and will bisect leading to it once I confirm
> 
> 

[root@ibclient ~]# uname -a
Linux ibclient 4.9.0+ #1 SMP Tue Dec 20 10:06:26 EST 2016 x86_64 x86_64 x86_64 GNU/Linux

Yep, back into 4.9 and we are at 1MB I/O

Confirmed regression or behavior change in 4.9

while true; do dd if=/dev/sdw bs=4096k of=/dev/null iflag=direct 1>/dev/null 2>&1 ; done &

# DISK STATISTICS (/sec)
#                   <---------reads---------------><---------writes--------------><--------averages--------> Pct
#Time     Name       KBytes Merged  IOs Size  Wait  KBytes Merged  IOs Size  Wait  RWSize  QLen  Wait SvcTim Util
16:06:33 sdw         1405952      0 1373 1024     2       0      0    0    0     0    1024     3     2      0   98
16:06:34 sdw         1424384      0 1391 1024     2       0      0    0    0     0    1024     3     2      0   97
16:06:35 sdw         1462272      0 1428 1024     2       0      0    0    0     0    1024     3     2      0   97
16:06:36 sdw         1417216      0 1384 1024     2       0      0    0    0     0    1024     3     2      0   97
16:06:37 sdw         1380352      0 1348 1024     2       0      0    0    0     0    1024     3     2      0   98
16:06:38 sdw         1404928      0 1372 1024     2       0      0    0    0     0    1024     3     2      0   98
16:06:39 sdw         1458176      0 1424 1024     2       0      0    0    0     0    1024     3     2      0   97

I will try chase it down and be back with an update

Thanks
Laurence
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

  parent reply	other threads:[~2016-12-20 21:09 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-12-08  1:10 [PATCH, RFC 0/5] IB: Optimize DMA mapping Bart Van Assche
     [not found] ` <07c07529-4636-fafb-2598-7358d8a1460d-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
2016-12-08  1:10   ` [PATCH 1/5] treewide: constify most struct dma_map_ops Bart Van Assche
     [not found]     ` <f6b70724-772c-c17f-f1be-1681fab31228-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
2016-12-09 18:20       ` Christoph Hellwig
2016-12-08  1:10   ` [PATCH 2/5] misc: vop: Remove a cast Bart Van Assche
     [not found]     ` <6fff2450-6442-4539-47ff-67f04a593c06-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
2016-12-09 18:20       ` Christoph Hellwig
2016-12-08  1:11   ` [PATCH 3/5] Move dma_ops from archdata into struct device Bart Van Assche
2016-12-09 18:22     ` Christoph Hellwig
2016-12-09 19:13       ` David Woodhouse
2016-12-09 19:46         ` Bart Van Assche
2016-12-08  1:11   ` [PATCH 4/5] IB: Switch from struct ib_dma_mapping_ops to struct dma_mapping_ops Bart Van Assche
     [not found]     ` <25d066c2-59d7-2be7-dd56-e29e99b43620-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
2016-12-09 18:23       ` Christoph Hellwig
2016-12-09 18:24       ` Christoph Hellwig
     [not found]         ` <20161209182429.GF16622-jcswGhMUV9g@public.gmane.org>
2016-12-19 16:42           ` Dennis Dalessandro
     [not found]             ` <52e8398f-a146-721c-3b92-0892b4abbff8-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
2016-12-19 16:55               ` Bart Van Assche
     [not found]                 ` <1482166487.25336.10.camel-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
2016-12-20 19:33                   ` Testing latest linux-next 4.9 ib_srp and ib_srpt Laurence Oberman
     [not found]                     ` <1918919536.8196250.1482262406057.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-20 19:43                       ` Laurence Oberman
     [not found]                         ` <2052479881.8196880.1482263028727.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-20 20:44                           ` Laurence Oberman
     [not found]                             ` <1668746735.8200653.1482266682585.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-20 21:09                               ` Laurence Oberman [this message]
2016-12-21  3:31                               ` Laurence Oberman
     [not found]                                 ` <120559766.8215321.1482291094018.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-21  6:34                                   ` Laurence Oberman
     [not found]                                     ` <1928955380.8220327.1482302041315.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-21  8:08                                       ` Bart Van Assche
2016-12-22  2:10                                       ` Laurence Oberman
     [not found]                                         ` <1337539588.8422488.1482372617094.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-22  6:23                                           ` Christoph Hellwig
     [not found]                                             ` <20161222062321.GA30326-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
2016-12-22 13:17                                               ` Testing latest linux-next 4.9 ib_srp and ib_srpt sees I/O capped at 1MB and no merging Laurence Oberman
     [not found]                                                 ` <1661819060.8462293.1482412678092.JavaMail.zimbra-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-12-22 15:40                                                   ` Christoph Hellwig
     [not found]                                                     ` <20161222154049.GA4638-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
2016-12-22 17:59                                                       ` Laurence Oberman
2016-12-22 19:55                                                       ` block: add back plugging in __blkdev_direct_IO kbuild test robot
2016-12-22 19:54                                                     ` kbuild test robot
2016-12-08  1:12   ` [PATCH 5/5] treewide: Inline ib_dma_map_*() functions Bart Van Assche
     [not found]     ` <9bdf696e-ec64-d60e-3d7e-7ad5b3000d60-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org>
2016-12-09 18:23       ` Christoph Hellwig
2016-12-08  6:48   ` [PATCH, RFC 0/5] IB: Optimize DMA mapping Or Gerlitz
     [not found]     ` <CAJ3xEMi4HY9Ehp-V4rP5UieAk=GjAu0X4uEnP-yMDomcFDpHkA-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2016-12-08 16:51       ` Bart Van Assche

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=653622322.8201942.1482268177727.JavaMail.zimbra@redhat.com \
    --to=loberman-h+wxahxf7alqt0dzr+alfa@public.gmane.org \
    --cc=Bart.VanAssche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org \
    --cc=linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).