From: "Nicholas A. Bellinger" <nab@linux-iscsi.org>
To: linux-iscsi-target-dev@googlegroups.com
Cc: linux-scsi <linux-scsi@vger.kernel.org>,
Christoph Hellwig <hch@lst.de>,
James Bottomley <James.Bottomley@HansenPartnership.com>
Subject: Re: [PATCH 0/42 RESEND+NEW] Target updates for May 27
Date: Fri, 27 May 2011 15:15:40 -0700 [thread overview]
Message-ID: <1306534540.23461.203.camel@haakon2.linux-iscsi.org> (raw)
In-Reply-To: <1306523240-15543-1-git-send-email-agrover@redhat.com>
On Fri, 2011-05-27 at 12:06 -0700, Andy Grover wrote:
> Hi Nicholas,
>
> The first 24 are a resend (rebased to current lio-41 git) of patches
> from a couple weeks ago.
>
> The last 18 are new. They include some misc. cleanups, but the heart
> of what they do is to allocate the memory for data buffers in the
> fabric (and from fabric cxt), instead of the core. This saves two context
> switches per command. It also reduces code size, since we no longer need
> to synchronize between the two threads, and we have more direct knowledge
> about the layout of the data buffer.
>
> iscsi fabric does not rely on se_cmd->t_mem_list at all anymore. After some
> more cleanups of other fabrics, t_mem_list and struct se_mem will be an
> implementation detail of tcm core (as opposed to part of its API), which
> will allow further progress.
>
> The code has been tested and appears to work fine, and although I haven't
> done any performance measurements, I'd expect to see command servicing
> latencies improve a bit.
>
Hi Andy,
My apologies for the delayed on the last series.. I will have a look at
these updated bits ahead of next week and start merging these into
lio-core-2.6.git and will follow up with specific comments.
Also, I CC'ed the linux-scsi list here just to keep folks in the loop
for the future.
Btw, I am quiet excited about the iscsi-target conversion to drop direct
se_cmd->t_mem_list usage and some of the other improvements. Great
stuff. :)
Thank you!
--nab
> Thanks -- Regards -- Andy
>
>
> The following changes since commit 41cb785aaaeeaa2ee3532afe36b9c3ff3e3d94b5:
>
> Merge branch 'master' of /pub/scm/linux/kernel/git/torvalds/linux-2.6 into lio-4.1 (2011-05-27 07:57:12 +0000)
>
> are available in the git repository at:
>
> git://fedorapeople.org/home/fedora/grover/public_git/linux-2.6.git for-nab
>
> Andy Grover (42):
> target: Replace custom sg<->buf functions with lib funcs
> target: Simplify sector limiting code
> target: get_cdb should never return NULL
> target: Simplify transport_memcpy_se_mem_read_contig
> target: Use assignment rather than increment for t_task_cdbs
> target: Don't pass dma_size to generic_get_mem. Don't memset.
> target: Pass sg with type scatterlist in transport_map_sg_to_mem
> target: Move task_sg_num next to task_sg in struct se_task
> target: inline struct se_transport_task into struct se_task
> target: Rewrite transport_init_task_sg()
> target: Simplify transport_do_task_sg_chain()
> target: Change name & semantics of transport_get_sectors()
> target: Remove unused members of se_cmd
> target: Rename se_cmd.t_task_cdbs to t_task_list_num
> target: Fix some spelling
> target/iscsi: rearrange iscsi_session members to show lock coverage
> target/iscsi: Fix spelling of enum immediate_data_ret_table members
> target/iscsi: Remove iscsi_ooo_cmdsn.batch_count
> target: Remove unused members of iscsi_cmd and iscsi_queue_req
> target: Remove unused var from transport_generic_do_tmr
> target: Remove iscsi_session.cmdsn_outoforder flag
> target: Remove iscsi_session.ooo_cmdsn_count
> target: Simplify CmdSN sequencing
> target: Silence a warning in a printk in target_emulate_write_same
> target: Rename transport_generic_allocate_tasks to _process_cdb
> target: Rename transport_generic_handle_cdb to _new_cmd
> target: map_sg_to_mem: return sg_count in return value
> target/pscsi: Use min_t for sector limits
> target/pscsi: Unused param for pscsi_get_bio()
> target: Rename get_cdb_count to allocate_tasks
> target: Have iscsi fabric allocate its own buffers
> target/iscsi: Inline params from map_sg into set_iovec_ptrs
> target/iscsi: pass data_length by value to iscsit_calculate_map_segment
> target/iscsi: Don't pass lmap to iscsit_get_offset()
> target/iscsi: Rewrite iscsit_get_offset()
> target/iscsi: Remove unused function iscsit_send_async_msg
> target/iscsi: Do not use t_mem_list anymore
> target/iscsi: use pad_bytes in cmd over local vars
> target: Call transport_new_cmd instead of adding to cmd queue
> target: Remove fabric callback to allocate iovecs
> target/iscsi: remove unsolicited_data_comp completion
> target/file: Alloc iov[] off the stack
>
> drivers/infiniband/ulp/srpt/ib_srpt.c | 22 +-
> drivers/target/iscsi/iscsi_target.c | 894 +++++------------
> drivers/target/iscsi/iscsi_target_configfs.c | 35 +-
> drivers/target/iscsi/iscsi_target_core.h | 39 +-
> drivers/target/iscsi/iscsi_target_erl1.c | 24 +-
> drivers/target/iscsi/iscsi_target_erl2.c | 1 -
> drivers/target/iscsi/iscsi_target_tmr.c | 4 +-
> drivers/target/iscsi/iscsi_target_util.c | 257 ++---
> drivers/target/iscsi/iscsi_target_util.h | 42 +-
> drivers/target/loopback/tcm_loop.c | 17 +-
> drivers/target/target_core_alua.c | 4 +-
> drivers/target/target_core_cdb.c | 41 +-
> drivers/target/target_core_device.c | 6 +-
> drivers/target/target_core_file.c | 27 +-
> drivers/target/target_core_iblock.c | 4 +-
> drivers/target/target_core_pr.c | 22 +-
> drivers/target/target_core_pscsi.c | 20 +-
> drivers/target/target_core_rd.c | 8 +-
> drivers/target/target_core_tmr.c | 56 +-
> drivers/target/target_core_transport.c | 1213 ++++++++++-------------
> drivers/target/target_core_ua.c | 2 +-
> drivers/target/tcm_fc/tfc_cmd.c | 18 +-
> drivers/target/tcm_fc/tfc_io.c | 21 +-
> drivers/target/tcm_qla2xxx/tcm_qla2xxx_fabric.c | 34 +-
> drivers/target/tcm_vhost/tcm_vhost_fabric.c | 2 +-
> drivers/target/tcm_vhost/tcm_vhost_scsi.c | 2 +-
> include/target/target_core_base.h | 105 +--
> include/target/target_core_fabric_ops.h | 5 -
> include/target/target_core_transport.h | 6 +-
> 29 files changed, 1071 insertions(+), 1860 deletions(-)
>
>
next parent reply other threads:[~2011-05-27 22:23 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <1306523240-15543-1-git-send-email-agrover@redhat.com>
2011-05-27 22:15 ` Nicholas A. Bellinger [this message]
2011-05-27 22:35 ` [PATCH 0/42 RESEND+NEW] Target updates for May 27 Christoph Hellwig
2011-05-27 23:39 ` Nicholas A. Bellinger
2011-05-28 7:43 ` Christoph Hellwig
2011-05-28 19:09 ` Nicholas A. Bellinger
2011-05-29 2:23 ` Andy Grover
2011-05-30 12:56 ` Christoph Hellwig
[not found] ` <1306523240-15543-11-git-send-email-agrover@redhat.com>
2011-05-31 7:08 ` [PATCH 10/42] target: Rewrite transport_init_task_sg() Nicholas A. Bellinger
2011-05-31 21:04 ` Andy Grover
2011-05-31 21:08 ` Christoph Hellwig
2011-06-01 0:33 ` Nicholas A. Bellinger
[not found] ` <1306523240-15543-38-git-send-email-agrover@redhat.com>
2011-05-31 9:00 ` [PATCH 37/42] target/iscsi: Do not use t_mem_list anymore Nicholas A. Bellinger
[not found] ` <1306523240-15543-40-git-send-email-agrover@redhat.com>
2011-05-31 9:32 ` [PATCH 39/42] target: Call transport_new_cmd instead of adding to cmd queue Nicholas A. Bellinger
2011-05-31 9:48 ` Christoph Hellwig
2011-05-31 10:10 ` Nicholas A. Bellinger
2011-05-31 10:22 ` Christoph Hellwig
2011-05-31 11:22 ` Nicholas A. Bellinger
2011-05-31 10:17 ` Christoph Hellwig
2011-05-31 11:18 ` Nicholas A. Bellinger
2011-06-01 4:09 ` Christoph Hellwig
2011-06-04 2:33 ` Nicholas A. Bellinger
2011-06-04 14:18 ` Christoph Hellwig
[not found] ` <1306523240-15543-42-git-send-email-agrover@redhat.com>
2011-05-31 9:58 ` [PATCH 41/42] target/iscsi: remove unsolicited_data_comp completion Nicholas A. Bellinger
[not found] ` <1306523240-15543-43-git-send-email-agrover@redhat.com>
2011-05-31 10:59 ` [PATCH 42/42] target/file: Alloc iov[] off the stack Nicholas A. Bellinger
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=1306534540.23461.203.camel@haakon2.linux-iscsi.org \
--to=nab@linux-iscsi.org \
--cc=James.Bottomley@HansenPartnership.com \
--cc=hch@lst.de \
--cc=linux-iscsi-target-dev@googlegroups.com \
--cc=linux-scsi@vger.kernel.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