linux-kernel - Re: [PATCH v3 char-misc-next] misc: microchip: pci1xxxx: Add OTP/EEPROM driver for the pci1xxxx switch

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: <Y+E9iRWyxqOjO4UX@kroah.com>
Date:   Mon, 6 Feb 2023 18:48:57 +0100
From:   Greg KH <gregkh@...uxfoundation.org>
To:     Tharunkumar.Pasumarthi@...rochip.com
Cc:     Kumaravel.Thiagarajan@...rochip.com, arnd@...db.de,
        linux-kernel@...r.kernel.org, linux-gpio@...r.kernel.org,
        UNGLinuxDriver@...rochip.com, lkp@...el.com
Subject: Re: [PATCH v3 char-misc-next] misc: microchip: pci1xxxx: Add
 OTP/EEPROM driver for the pci1xxxx switch

On Mon, Feb 06, 2023 at 05:03:57PM +0000, Tharunkumar.Pasumarthi@...rochip.com wrote:
> > From: Greg KH <gregkh@...uxfoundation.org>
> > Sent: Monday, February 6, 2023 3:12 PM
> > To: Tharunkumar Pasumarthi - I67821
> > <Tharunkumar.Pasumarthi@...rochip.com>
> > > > > +     delay = msecs_to_jiffies(OTP_RW_TIMEOUT_MILLISECONDS);
> > > > > +     j0 = jiffies;
> > > > > +     j1 = j0 + delay;
> > > >
> > > > Are you sure this math works out?  Please use the jiffies math
> > > > functions instead so you can handle wrapping properly.
> > >
> > > You suggest using any existing APIs to add jiffies to handle wrapping?
> > > I am not able to find any such API. Can you please point out API name.
> > 
> > I have no context here for what you are doing with the math in jiffies, but
> > that is usually a very odd thing.  If you use the normal timer functions, you
> > will not have to worry about the wrapping as it is handled for you, right?
> 
> Hi Greg,
> Our objective with that part of code is to wait for OTP_STATUS_BUSY_BIT 
> to be cleared in the hardware, but make sure that we wait for a maximum 
> of OTP_RW_TIMEOUT_MILLISECONDS ms in the driver and not beyond that. 

How exact much do you mean by "not beyond that"?

> This logic is not possible with normal timer functions. Timer wrapping logic is 
> handled by time_before API internally. So, there does not seem to be issue with
> existing code. 

I have no context here as to what the code actually looked like anymore,
so I do not know what is going on anymore, sorry.

greg k-h