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: <ZRwbKRnnKY/tDqCF@yury-ThinkPad>
Date:   Tue, 3 Oct 2023 06:46:17 -0700
From:   Yury Norov <yury.norov@...il.com>
To:     Paolo Abeni <pabeni@...hat.com>
Cc:     linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
        linux-rdma@...r.kernel.org, Tariq Toukan <ttoukan.linux@...il.com>,
        Valentin Schneider <vschneid@...hat.com>,
        Maher Sanalla <msanalla@...dia.com>,
        Ingo Molnar <mingo@...nel.org>, Mel Gorman <mgorman@...e.de>,
        Saeed Mahameed <saeedm@...dia.com>,
        Leon Romanovsky <leon@...nel.org>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Juri Lelli <juri.lelli@...hat.com>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        Dietmar Eggemann <dietmar.eggemann@....com>,
        Steven Rostedt <rostedt@...dmis.org>,
        Ben Segall <bsegall@...gle.com>,
        Daniel Bristot de Oliveira <bristot@...hat.com>,
        Pawel Chmielewski <pawel.chmielewski@...el.com>,
        Jacob Keller <jacob.e.keller@...el.com>,
        Yury Norov <ynorov@...dia.com>
Subject: Re: [PATCH 1/4] net: mellanox: drop mlx5_cpumask_default_spread()

On Tue, Oct 03, 2023 at 12:04:01PM +0200, Paolo Abeni wrote:
> On Sun, 2023-09-24 at 19:05 -0700, Yury Norov wrote:
> > The function duplicates existing cpumask_local_spread(), and it's O(N),
> > while cpumask_local_spread() implementation is based on bsearch, and
> > thus is O(log n), so drop mlx5_cpumask_default_spread() and use generic
> > cpumask_local_spread().
> > 
> > Signed-off-by: Yury Norov <yury.norov@...il.com>
> > Signed-off-by: Yury Norov <ynorov@...dia.com>
> > ---
> >  drivers/net/ethernet/mellanox/mlx5/core/eq.c | 28 ++------------------
> >  1 file changed, 2 insertions(+), 26 deletions(-)
> > 
> > diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
> > index ea0405e0a43f..bd9f857cc52d 100644
> > --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
> > +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
> > @@ -828,30 +828,6 @@ static void comp_irq_release_pci(struct mlx5_core_dev *dev, u16 vecidx)
> >  	mlx5_irq_release_vector(irq);
> >  }
> >  
> > -static int mlx5_cpumask_default_spread(int numa_node, int index)
> > -{
> > -	const struct cpumask *prev = cpu_none_mask;
> > -	const struct cpumask *mask;
> > -	int found_cpu = 0;
> > -	int i = 0;
> > -	int cpu;
> > -
> > -	rcu_read_lock();
> > -	for_each_numa_hop_mask(mask, numa_node) {
> > -		for_each_cpu_andnot(cpu, mask, prev) {
> > -			if (i++ == index) {
> > -				found_cpu = cpu;
> > -				goto spread_done;
> > -			}
> > -		}
> > -		prev = mask;
> > -	}
> > -
> > -spread_done:
> > -	rcu_read_unlock();
> > -	return found_cpu;
> > -}
> > -
> >  static struct cpu_rmap *mlx5_eq_table_get_pci_rmap(struct mlx5_core_dev *dev)
> >  {
> >  #ifdef CONFIG_RFS_ACCEL
> > @@ -873,7 +849,7 @@ static int comp_irq_request_pci(struct mlx5_core_dev *dev, u16 vecidx)
> >  	int cpu;
> >  
> >  	rmap = mlx5_eq_table_get_pci_rmap(dev);
> > -	cpu = mlx5_cpumask_default_spread(dev->priv.numa_node, vecidx);
> > +	cpu = cpumask_local_spread(vecidx, dev->priv.numa_node);
> >  	irq = mlx5_irq_request_vector(dev, cpu, vecidx, &rmap);
> >  	if (IS_ERR(irq))
> >  		return PTR_ERR(irq);
> > @@ -1125,7 +1101,7 @@ int mlx5_comp_vector_get_cpu(struct mlx5_core_dev *dev, int vector)
> >  	if (mask)
> >  		cpu = cpumask_first(mask);
> >  	else
> > -		cpu = mlx5_cpumask_default_spread(dev->priv.numa_node, vector);
> > +		cpu = cpumask_local_spread(vector, dev->priv.numa_node);
> >  
> >  	return cpu;
> >  }
> 
> It looks like this series is going to cause some later conflicts
> regardless of the target tree. I think the whole series could go via
> the net-next tree, am I missing any relevant point?

Hi Paolo,

Can you elaborate on the conflicts you see? For me it applies cleanly
on current master, and with some 3-way merging on latest -next...

Thanks,
Yury

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ