linux-kernel - Re: [PATCH 0/4] btrfs: Convert kmaps to core page calls

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

Message-ID: <20210209215229.GC2975576@iweiny-DESK2.sc.intel.com>
Date:   Tue, 9 Feb 2021 13:52:29 -0800
From:   Ira Weiny <ira.weiny@...el.com>
To:     Andrew Morton <akpm@...ux-foundation.org>
Cc:     dsterba@...e.cz, clm@...com, josef@...icpanda.com,
        dsterba@...e.com, linux-kernel@...r.kernel.org,
        linux-fsdevel@...r.kernel.org, Matthew Wilcox <willy@...radead.org>
Subject: Re: [PATCH 0/4] btrfs: Convert kmaps to core page calls

On Tue, Feb 09, 2021 at 01:11:03PM -0800, Andrew Morton wrote:
> > > 
> > > It would be best to merge [1/4] via the btrfs tree.  Please add my
> > > 
> > > Acked-by: Andrew Morton <akpm@...ux-foundation.org>
> > > 
> > > 
> > > Although I think it would be better if [1/4] merely did the code
> > > movement.  Adding those BUG_ON()s is a semantic/functional change and
> > > really shouldn't be bound up with the other things this patch series
> > > does.
> > 
> > I proposed this too and was told 'no'...
> > 
> > <quote>
> > If we put in into a separate patch, someone will suggest backing out the
> > patch which tells us that there's a problem.
> > </quote>
> > 	-- https://lore.kernel.org/lkml/20201209201415.GT7338@casper.infradead.org/
> 
> Yeah, no, please let's not do this.  Bundling an offtopic change into
> [1/4] then making three more patches dependent on the ontopic parts of
> [1/4] is just rude.
> 
> I think the case for adding the BUG_ONs can be clearly made.  And that
> case should at least have been clearly made in the [1/4] changelog!
> 
> (Although I expect VM_BUG_ON() would be better - will give us sufficient
> coverage without the overall impact.)

I'm ok with VM_BUG_ON()

> 
> Let's please queue this up separately.

Ok can I retain your Ack on the move part of the patch?  Note that it does
change kmap_atomic() to kmap_local_page() currently.

Would you prefer a separate change for that as well?

Ira

PS really CC'ing Matthew now...