Re: [PATCH 08/22] qla2xxx: Refactore target code to remove symbol dependency

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Christoph,

Qlogic was asked to support other Target Stack (SC ST) using the same qla2xxx.ko upstream driver.  I had mentioned this during the last 2015 LSF get together.  Hopefully, that gives a better context.  The understanding that came away was “Qlogic do what is needed to provide this support without creating any dead code”.

As for the cleanup patch(es), we fail to mention that conversation.  The patch attempts to remove any knowledge of TCM’s specific symbol and try to formalize the interface.  Qla2xxx driver will behave in a more generic manner when it comes to Target Mode.  All TCM knowledge will be confined to the tcm_qla2xxx driver before communicating with QLA driver.   The same goes for other target stack.

The patches look like its copying a lot of fields without doing a whole lot, instead the patches are translating the request from tcm_qla2xxx to qla2xxx.

As for T10-PI change in this series, the goal were the same : 1) remove TCM knowledge/cleanup, 2) additional bug fixes were needed while testing with other stacks. 

Regards,
Quinn Tran

-----Original Message-----
From: <linux-scsi-owner@xxxxxxxxxxxxxxx> on behalf of "Madhani, Himanshu" <Himanshu.Madhani@xxxxxxxxxx>
Date: Thursday, December 15, 2016 at 11:29 AM
To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Cc: target-devel <target-devel@xxxxxxxxxxxxxxx>, Nicholas Bellinger <nab@xxxxxxxxxxxxxxx>, "Malavali, Giridhar" <Giridhar.Malavali@xxxxxxxxxx>, "linux-scsi@xxxxxxxxxxxxxxx" <linux-scsi@xxxxxxxxxxxxxxx>
Subject: Re: [PATCH 08/22] qla2xxx: Refactore target code to remove symbol dependency
Resent-From: <quinn.tran@xxxxxxxxxx>

    Hi Christoph, 
    
    
    
    On 12/14/16, 1:18 PM, "Christoph Hellwig" <hch@xxxxxxxxxxxxx> wrote:
    
    > - the new qla2x00_free_fcport is entirely pointless, please drop
    >   that part of the patch (and even if it wasn't pointless it should
    >   have been a patch on it's own)
    > - please use struct names and not typedefs for your new structures
    > - pretty much avery single items in your list should be a separate
    >   patch.  And some of them are actively counterproductive and should
    >   be dropped:
    >
    >        - Remove direct access of scsi_status field in se_cmd
    >	- Remove se_cmd from qlt_do_ctio_completion
    >	- Remove se_cmd access in srr code section
    >	- Move se_cmd struct outside of qla_tgt_cmd/qla_tgt_mgmt_cmd.
    
    We combined the patches which includes fixes, enhancements and cleanups to 
    Support multiple target stacks, reviews from customer, and to improve code
    maintainability. This patch tries to organize code logically which is spread
    across qla_target and tcm_qla2xxx. We will split this patch into multiple 
    subset and resubmit the series. 
    
    Thanks, 
    - Himanshu
    >
    ?�{.n�+�������+%��lzwm��b�맲��r��zX��??�(��?��ܨ}���Ơz�&j:+v���?����zZ+��+zf���h���~����i���z�?�w���?����&�)ߢ?f

��.n��������+%������w��{.n�����{������ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f




[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [SCSI Target Devel]     [Linux SCSI Target Infrastructure]     [Kernel Newbies]     [IDE]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux ATA RAID]     [Linux IIO]     [Samba]     [Device Mapper]

  Powered by Linux