linux-kernel - Re: [PATCH v1 1/1] rxe: calculate inline data size based on requested values

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: <20191029194437.GI6128@ziepe.ca>
Date:   Tue, 29 Oct 2019 16:44:38 -0300
From:   Jason Gunthorpe <jgg@...pe.ca>
To:     Rao Shoaib <rao.shoaib@...cle.com>
Cc:     monis@...lanox.com, dledford@...hat.com, sean.hefty@...el.com,
        hal.rosenstock@...il.com, linux-rdma@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1 1/1] rxe: calculate inline data size based on
 requested values

On Tue, Oct 29, 2019 at 12:31:03PM -0700, Rao Shoaib wrote:

> > > @@ -81,6 +80,7 @@ enum rxe_device_param {
> > >   					| IB_DEVICE_MEM_MGT_EXTENSIONS,
> > >   	RXE_MAX_SGE			= 32,
> > >   	RXE_MAX_SGE_RD			= 32,
> > > +	RXE_MAX_INLINE_DATA		= RXE_MAX_SGE * sizeof(struct ib_sge),
> > >   	RXE_MAX_CQ			= 16384,
> > >   	RXE_MAX_LOG_CQE			= 15,
> > >   	RXE_MAX_MR			= 2 * 1024,
> > Increasing RXE_MAX_INLINE_DATA to match the WQE size limited the
> > MAX_SGE. IMHO this is done in a hacky way, instead we should define a
> > maximim WQE size and from there derive the MAX_INLINE_DATA and MAX_SGE
> > limitations.
> There was already RXE_MAX_SGE defined so I did not define MAX_WQE. If that
> is what is preference I can submit a patch with that. What is a good value
> for MAX_WQE?

I would arrange it so that RXE_MAX_SGE doesn't change

> > Also don't double initialize qp->sq.max_inline in the same function,
> > and there is no need for the temporary 'inline_size'
> 
> I used a separate variable as I would have to repeat the calculation twice.
> I do not understand your comment about double initialization, can you please
> clarify that for me.

Assign it to qp->sq.max_inline and then read it to get the init

Look above in the function, there is already an assignment to
qp->sq.max_inline

Jason