linux-kernel - RE: [PATCH v3 02/18] NTB: Add NTB hardware abstraction layer

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <40F65EF2B5E2254199711F58E3ACB84D830FFF99@MX104CL02.corp.emc.com>
Date:	Tue, 9 Jun 2015 17:32:01 +0000
From:	"Hubbe, Allen" <Allen.Hubbe@....com>
To:	Bjorn Helgaas <bhelgaas@...gle.com>
CC:	"linux-ntb@...glegroups.com" <linux-ntb@...glegroups.com>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
	Jon Mason <jdmason@...zu.us>, Dave Jiang <dave.jiang@...el.com>
Subject: RE: [PATCH v3 02/18] NTB: Add NTB hardware abstraction layer

From: Bjorn Helgaas [mailto:bhelgaas@...gle.com]
> > It was pointed out in v1 that this one patch is over 200KB.  Please
> > accept my appologies for not breaking this down further.  The affected
> > components in this patch had interdependencies, which makes it
> difficult
> > to contain the changes.  This change separates the components, so
> future
> > changes can be better contained.
> >
...
> 
> I'm not really a reviewer (I'm only commenting on obvious English and
> spelling issues), so maybe the real reviewers can handle this.  But if I
> were reviewing the code, this patch would be far too large for me to
> make
> sense of.
> 
> And it makes bisection and reversion significantly less useful.
> 
> Bjorn

I'm not happy about it either.  I did make a few attempts to split this, so I can say at least that it will not be easy.

I really can't change ntb_transport and ntb_hw_intel separately.  They call each other, so splitting one driver to use the abstraction layer requires the other driver to change at the same time, or the other driver breaks.  Unfortunately, splitting these drivers is the large bulk of the patch.

I also tried to just add ntb.h, ntb.c separately, before spliting ntb_hw_intel and ntb_transport.  I forget exactly why I backed out of that change a couple weeks ago.  I think it has something to do with the wonky make targets in drivers/ntb after the first patch.  I'll try once again to split it this way.