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: <20250319213327.2bf94bc1@pumpkin>
Date: Wed, 19 Mar 2025 21:33:27 +0000
From: David Laight <david.laight.linux@...il.com>
To: <shao.mingyin@....com.cn>
Cc: <hch@....de>, <sagi@...mberg.me>, <kch@...dia.com>,
 <linux-nvme@...ts.infradead.org>, <linux-kernel@...r.kernel.org>,
 <yang.tao172@....com.cn>, <yang.yang29@....com.cn>, <xu.xin16@....com.cn>
Subject: Re: [PATCH] nvme: replace max(a, min(b, c)) by clamp(b, a, c)

On Wed, 19 Mar 2025 21:00:32 +0000
David Laight <david.laight.linux@...il.com> wrote:

> On Mon, 17 Mar 2025 15:39:09 +0800 (CST)
> <shao.mingyin@....com.cn> wrote:
> 
> > From: LiHaoran <li.haoran7@....com.cn>
> > 
> > This patch replaces max(a, min(b, c)) by clamp(b, a, c) in the nvme
> > driver. This improves the readability.
> > 
> > Signed-off-by: LiHaoran <li.haoran7@....com.cn>
> > Cc: ShaoMingyin <shao.mingyin@....com.cn>
> > ---
> >  drivers/nvme/target/nvmet.h | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/nvme/target/nvmet.h b/drivers/nvme/target/nvmet.h
> > index fcf4f460dc9a..30804b0ca66e 100644
> > --- a/drivers/nvme/target/nvmet.h
> > +++ b/drivers/nvme/target/nvmet.h
> > @@ -819,7 +819,7 @@ static inline u8 nvmet_cc_iocqes(u32 cc)
> >  /* Convert a 32-bit number to a 16-bit 0's based number */
> >  static inline __le16 to0based(u32 a)
> >  {
> > -	return cpu_to_le16(max(1U, min(1U << 16, a)) - 1);
> > +	return cpu_to_le16(clamp(1U << 16, 1U, a) - 1);  
> 
> Swap the arguments into a sane order - clamp(a, 1, 1 << 16)

To clarify further it is clamp(val, lo, hi) and is only well defined if 'lo <= hi'.
In particular the order of the comparisons is not defined.
It might be val > hi ? hi : val < lo ? lo : val.
Which won't give the value you expect when 'a' is zero.

> 
> 	David
> 
> >  }
> > 
> >  static inline bool nvmet_ns_has_pi(struct nvmet_ns *ns)  
> 


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ