public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Andreas Dilger <adilger@turbolabs.com>
To: Christoph Hellwig <hch@caldera.de>, linux-kernel@vger.kernel.org
Subject: Re: 2.5.1-pre2 does not compile
Date: Wed, 28 Nov 2001 10:27:15 -0700	[thread overview]
Message-ID: <20011128102715.R730@lynx.no> (raw)
In-Reply-To: <200111272209.fARM9tk18991@ns.caldera.de> <Pine.LNX.4.33.0111271628430.1629-100000@penguin.transmeta.com> <20011128135508.A21418@caldera.de> <20011128092600.Q730@lynx.no> <20011128174250.A17582@caldera.de>
In-Reply-To: <20011128174250.A17582@caldera.de>; from hch@caldera.de on Wed, Nov 28, 2001 at 05:42:50PM +0100

On Nov 28, 2001  17:42 +0100, Christoph Hellwig wrote:
> On Wed, Nov 28, 2001 at 09:26:00AM -0700, Andreas Dilger wrote:
> > What would be nice in the case of drivers that don't use the new error
> > handling code is to add something like:
> > 
> > #warning "Uses obsolete SCSI error code, see Documentation/2.5/scsi-error.txt"
> > 
> > for a hint as to the reason why it no longer compiles, and a short guide
> > on how to update the drivers.
> 
> I already thought about that - as the old error handling code is selected
> by setting a member in a struct to '1' I don't see any easy way to do so...

I'm thinking about a compile-time #warning/#error, that at least tells an
interested party where to start looking for how to fix this.  It could be
anywhere in the affected driver source file (preferrably at the top, so it
is hit before the compiler exits because of too many other errors).  It
would just be a matter of cut-n-paste a single line into every file which
previously used the old error-handling code.

I presume that the nature of this change will either cause compile errors
anyways (because of missing struct definitions), or linker errors (because
of missing functions).  I think that having a #warning at compile time will
at least avoid a whole bunch of "this driver doesn't compile, help me"
emails, and may also help someone to actually update the driver rather than
giving up because they have no idea where to begin looking for the fix.

Actually, what would also be nice is to include the kernel version where
this change happened, so that interested parties could also see what changes
where necessary to bring about this fix.  In this case (removing the old
error handling support), it probably doesn't help much, but in Jens' BIO
changes it _would_ be very helpful to know when a major change was made
so potential fixers can "follow along" with the changes rather than having
to reverse-engineer it 50 releases from now (along with the 33 other major
changes that have been made along the way).  Even now, with the BIO +
SCSI error handling changes there are a lot of SCSI drivers with problems
and it won't be "obvious" how to fix them later on.

Cheers, Andreas

PS - are you sure the old error handling code was selected with a '1' and
     not with a '0' for "use_new_eh_code"?
--
Andreas Dilger
http://sourceforge.net/projects/ext2resize/
http://www-mddsp.enel.ucalgary.ca/People/adilger/


  parent reply	other threads:[~2001-11-28 17:27 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-11-27 20:44 2.5.1-pre2 does not compile f5ibh
2001-11-27 20:50 ` Linus Torvalds
2001-11-27 22:02   ` Paul Mackerras
2001-11-28  1:04     ` Linus Torvalds
2001-11-28  1:34       ` Block I/O Enchancements, 2.5.1-pre2 Jeff V. Merkey
2001-11-28  1:45         ` Jeff Garzik
2001-11-28  1:55           ` Jeff V. Merkey
2001-11-28  2:32             ` Jeff Garzik
2001-11-28  3:38               ` Linus Torvalds
2001-11-30  2:19                 ` Daniel Phillips
2001-11-30 12:21                   ` Hans Reiser
     [not found]                     ` <15367.32910.275973.287742@laputa.namesys.com>
2001-12-01  9:31                       ` [reiserfs-dev] " Hans Reiser
2001-11-28 10:17               ` Martin Dalecki
2001-11-28 13:35         ` Jens Axboe
2001-11-28 17:29           ` Jeff Merkey
2001-11-28  6:58       ` 2.5.1-pre2 does not compile Jens Axboe
2001-11-28 12:20       ` bio write-up (was: Re: 2.5.1-pre2 does not compile) Jens Axboe
2001-11-28 23:31       ` 2.5.1-pre2 bio offset by one error in VIA IDE Anton Altaparmakov
2001-11-28 23:55         ` Andreas Dilger
2001-11-29  1:07         ` Anton Altaparmakov
2001-11-30  1:53     ` 2.5.1-pre2 does not compile Daniel Phillips
2001-11-27 22:09   ` Christoph Hellwig
2001-11-27 22:22     ` onboard ethernet/sound on Soyo SY-K7V? Dax Kelson
2001-11-27 22:47       ` François Cami
2001-11-27 22:57       ` Jeff Garzik
2001-11-27 22:29     ` 2.5.1-pre2 does not compile Robert Love
2001-11-28  0:29     ` Linus Torvalds
2001-11-28 12:55       ` Christoph Hellwig
2001-11-28 16:26         ` Andreas Dilger
2001-11-28 16:42           ` Christoph Hellwig
2001-11-28 16:39             ` Martin Dalecki
2001-11-28 17:27             ` Andreas Dilger [this message]
2001-11-28  0:40   ` Andre Hedrick
  -- strict thread matches above, loose matches on Subject: below --
2001-11-27 21:02 Wayne.Brown

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=20011128102715.R730@lynx.no \
    --to=adilger@turbolabs.com \
    --cc=hch@caldera.de \
    --cc=linux-kernel@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