[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20180613005807.GB2265@hle-laptop.local>
Date: Tue, 12 Jun 2018 20:58:07 -0400
From: Hugo Lefeuvre <hle@....eu.com>
To: Dan Carpenter <dan.carpenter@...cle.com>
Cc: devel@...verdev.osuosl.org,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
kernelnewbies@...nelnewbies.org, linux-kernel@...r.kernel.org,
Valentin Vidic <Valentin.Vidic@...Net.hr>
Subject: Re: [PATCH v2] staging: pi433: add mutex fixing concurrency issues.
> > case PI433_IOC_WR_TX_CFG:
> > if (copy_from_user(&instance->tx_cfg, argp,
> > sizeof(struct pi433_tx_cfg)))
> > return -EFAULT;
> > break;
>
> Btw, it looks so wrong to me that we copy partial data to
> &instance->tx_cfg... I'd really prefer copying it to a tmp buffer and
> then verifying it's corrent then memcpy()ing it to &instance->tx_cfg.
AFAIK this piece of code is not supposed to check passed tx config.
These checks are realised later by rf69_set_tx_cfg (called by
pi433_tx_thread) after the config has been written to the fifo by
pi433_write.
What kind of checks do you want to perform exactly ?
But, right, I prefer the idea of the temporary buffer too, and seeing the
rest of kernel code it seems to be the usual way to go.
Regards,
Hugo
--
Hugo Lefeuvre (hle) | www.owl.eu.com
4096/ 9C4F C8BF A4B0 8FC5 48EB 56B8 1962 765B B9A8 BACA
Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)
Powered by blists - more mailing lists