DPDK patches and discussions
 help / color / mirror / Atom feed
From: Stephen Hemminger <stephen@networkplumber.org>
To: Slawomir Mrozowicz <slawomirx.mrozowicz@intel.com>,
	Declan Doherty <declan.doherty@intel.com>q
Cc: dev@dpdk.org
Subject: [dpdk-dev] bugs and glitches in rte_cryptodev_devices_get
Date: Wed, 1 Feb 2017 08:53:50 -0800	[thread overview]
Message-ID: <20170201085321.7168505a@xeon-e3> (raw)

The function rte_cryptodev_devices_get has several issues. I was just going to
fix it, but think it need to be explained.
 
One potentially serious one (reported by coverity) is:

*** CID 141067:    (BAD_COMPARE)
/lib/librte_cryptodev/rte_cryptodev.c: 503 in rte_cryptodev_devices_get()
497     				&& (*devs + i)->attached ==
498     						RTE_CRYPTODEV_ATTACHED) {
499     
500     			dev = (*devs + i)->device;
501     
502     			if (dev)
>>>     CID 141067:    (BAD_COMPARE)
>>>     Truncating the result of "strncmp" to "unsigned char" may cause it to be misinterpreted as 0. Note that "strncmp" may return an integer besides -1, 0, or 1.  
503     				cmp = strncmp(dev->driver->name,
504     						dev_name,
505     						strlen(dev_name));
506     			else
507     				cmp = strncmp((*devs + i)->data->name,
508     						dev_name,
/lib/librte_cryptodev/rte_cryptodev.c: 507 in rte_cryptodev_devices_get()
501     
502     			if (dev)
503     				cmp = strncmp(dev->driver->name,
504     						dev_name,
505     						strlen(dev_name));
506     			else
>>>     CID 141067:    (BAD_COMPARE)
>>>     Truncating the result of "strncmp" to "unsigned char" may cause it to be misinterpreted as 0. Note that "strncmp" may return an integer besides -1, 0, or 1.  
507     				cmp = strncmp((*devs + i)->data->name,
508     						dev_name,
509     						strlen(dev_name));
510     
511     			if (cmp == 0)
512     				devices[count++] = (*devs + i)->data->dev_id;


But also:

1. Incorrect function signature:
    * function returns int but never a negative value. should be unsigned.
    * devices argument is not modified should be const.

2. Original ABI seems short sighted with limit of 256 cryptodevs
    * this seems like 8 bit mindset,  should really use unsigned int instead
      of uint8_t for number of devices.

3. Wacky indention of the if statement.

4. Make variables local to the block they are used (cmp, dev)

5. Use array instead of pointer:
     ie. instead of *devs + i use devs[i]


The overall code in question is:


int
rte_cryptodev_devices_get(const char *dev_name, uint8_t *devices,
	uint8_t nb_devices)
{
	uint8_t i, cmp, count = 0;
	struct rte_cryptodev **devs = &rte_cryptodev_globals->devs;
	struct rte_device *dev;

	for (i = 0; i < rte_cryptodev_globals->max_devs && count < nb_devices;
			i++) {

		if ((*devs + i)
				&& (*devs + i)->attached ==
						RTE_CRYPTODEV_ATTACHED) {

			dev = (*devs + i)->device;

			if (dev)
				cmp = strncmp(dev->driver->name,
						dev_name,
						strlen(dev_name));
			else
				cmp = strncmp((*devs + i)->data->name,
						dev_name,
						strlen(dev_name));

			if (cmp == 0)
				devices[count++] = (*devs + i)->data->dev_id;
		}
	}

	return count;
}

Please fix it.

             reply	other threads:[~2017-02-01 16:53 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-02-01 16:53 Stephen Hemminger [this message]
2017-02-02 13:55 ` Mrozowicz, SlawomirX
2017-02-03 12:26 ` Mrozowicz, SlawomirX

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170201085321.7168505a@xeon-e3 \
    --to=stephen@networkplumber.org \
    --cc=declan.doherty@intel.com \
    --cc=dev@dpdk.org \
    --cc=slawomirx.mrozowicz@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).