netdev - Re: [net 3/6] net/mlx5: FPGA, return -EINVAL if size is zero

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]

Date:   Fri, 10 Nov 2017 15:37:57 +0900
From:   Saeed Mahameed <saeedm@...lanox.com>
To:     Or Gerlitz <gerlitz.or@...il.com>
Cc:     Kamal Heib <kamalh@...lanox.com>,
        Linux Netdev List <netdev@...r.kernel.org>,
        David Miller <davem@...emloft.net>,
        Yevgeny Kliteynik <kliteyn@...lanox.com>
Subject: Re: [net 3/6] net/mlx5: FPGA, return -EINVAL if size is zero

On Fri, 2017-11-10 at 15:23 +0900, Or Gerlitz wrote:
> On Fri, Nov 10, 2017 at 3:13 PM, Saeed Mahameed <saeedm@...lanox.com>
> wrote:
> > On Thu, 2017-11-09 at 18:12 +0900, Or Gerlitz wrote:
> > > On Thu, Nov 9, 2017 at 4:43 PM, Kamal Heib <kamalh@...lanox.com>
> > > wrote:
> > > > On Wed, 2017-11-08 at 23:13 +0900, Or Gerlitz wrote:
> > > > > On Wed, Nov 8, 2017 at 4:21 PM, Saeed Mahameed <saeedm@...lan
> > > > > ox.c
> > > > > om>
> > > > > wrote:
> > > > > > From: Kamal Heib <kamalh@...lanox.com>
> > > > > > 
> > > > > > In the current code, if a size of zero is passed to
> > > > > > mlx5_fpga_mem_{read|write}_i2c() functions the "err"
> > > > > 
> > > > > Don't we need to fix the call site where zero size is
> > > > > provided
> > > > > and
> > > > > not
> > > > > in called function?
> > > > > 
> > > > 
> > > > Isn't sending down a zero size a sign for a bug which we are
> > > > not
> > > > fixing?
> > > > > 
> > > > 
> > > > Both functions are called from an exported symbols. so I think
> > > > the
> > > > size
> > > > validation should be within this two functions just like the
> > > > case
> > > > of
> > > > checking that mdev isn't set.
> > > 
> > > mmm, I see exported to who exactly? how are they being called, by
> > > func pointer?
> > > can you point to the call sites?
> > 
> > Or, are you ok with this patch ? I would like to post V2 with the
> > reviewed-by tag fix.
> 
> The RB tag issue was on another patch.. for this patch I realized
> after talking
> to the author that it comes to fix a build warning. I would be happy
> if we can clarify
> that in the change log.


Ok I will drop this patch until the author provides the missing
information.