Re: [PATCH v3] staging: nrf24: add new driver for 2.4GHz radio transceiver

From: Marcin Ciupak
Date: Fri Oct 19 2018 - 17:35:48 EST


On Fri, Oct 19, 2018 at 09:58:01AM +0300, Dan Carpenter wrote:
> Hi Marcin,
>
> Thank you for the patch! Perhaps something to improve:
>
> url: https://github.com/0day-ci/linux/commits/Marcin-Ciupak/staging-nrf24-add-new-driver-for-2-4GHz-radio-transceiver/20181018-164230
>
> smatch warnings:
> drivers/staging/nrf24/nrf24_if.c:538 nrf24_create_pipe() error: 'p->dev' dereferencing possible ERR_PTR()
> drivers/staging/nrf24/nrf24_if.c:753 nrf24_probe() error: 'device' dereferencing possible ERR_PTR()
> drivers/staging/nrf24/nrf24_sysfs.c:118 plw_store() warn: impossible condition '(old < 0) => (0-255 < 0)'
>
> # https://github.com/0day-ci/linux/commit/f4952b4a228f4dc0149a814630c94e0e87a9f473
> git remote add linux-review https://github.com/0day-ci/linux
> git remote update linux-review
> git checkout f4952b4a228f4dc0149a814630c94e0e87a9f473
> vim +538 drivers/staging/nrf24/nrf24_if.c
>
> f4952b4a Marcin Ciupak 2018-10-17 503
> f4952b4a Marcin Ciupak 2018-10-17 504 static struct nrf24_pipe *nrf24_create_pipe(struct nrf24_device *device, int id)
> f4952b4a Marcin Ciupak 2018-10-17 505 {
> f4952b4a Marcin Ciupak 2018-10-17 506 int ret;
> f4952b4a Marcin Ciupak 2018-10-17 507 struct nrf24_pipe *p;
> f4952b4a Marcin Ciupak 2018-10-17 508
> f4952b4a Marcin Ciupak 2018-10-17 509 //sets flags to false as well
> f4952b4a Marcin Ciupak 2018-10-17 510 p = kzalloc(sizeof(*p), GFP_KERNEL);
> f4952b4a Marcin Ciupak 2018-10-17 511 if (!p) {
> f4952b4a Marcin Ciupak 2018-10-17 512 ret = -ENOMEM;
> f4952b4a Marcin Ciupak 2018-10-17 513 goto err_return;
> f4952b4a Marcin Ciupak 2018-10-17 514 }
> f4952b4a Marcin Ciupak 2018-10-17 515
> f4952b4a Marcin Ciupak 2018-10-17 516 ret = ida_simple_get(&nrf24_ida_pipe, 0, 0, GFP_KERNEL);
> f4952b4a Marcin Ciupak 2018-10-17 517 if (ret < 0) {
> f4952b4a Marcin Ciupak 2018-10-17 518 dev_err(&device->dev, "%s: get_minor failed\n", __func__);
> f4952b4a Marcin Ciupak 2018-10-17 519 goto err_free_mem;
> f4952b4a Marcin Ciupak 2018-10-17 520 }
> f4952b4a Marcin Ciupak 2018-10-17 521
> f4952b4a Marcin Ciupak 2018-10-17 522 p->devt = MKDEV(MAJOR(nrf24_dev), ret);
> f4952b4a Marcin Ciupak 2018-10-17 523 p->id = id;
> f4952b4a Marcin Ciupak 2018-10-17 524
> f4952b4a Marcin Ciupak 2018-10-17 525 INIT_KFIFO(p->rx_fifo);
> f4952b4a Marcin Ciupak 2018-10-17 526 init_waitqueue_head(&p->poll_wait_queue);
> f4952b4a Marcin Ciupak 2018-10-17 527
> f4952b4a Marcin Ciupak 2018-10-17 528 p->dev = device_create_with_groups(nrf24_class,
> f4952b4a Marcin Ciupak 2018-10-17 529 &device->dev,
> f4952b4a Marcin Ciupak 2018-10-17 530 p->devt,
> f4952b4a Marcin Ciupak 2018-10-17 531 p,
> f4952b4a Marcin Ciupak 2018-10-17 532 nrf24_pipe_groups,
> f4952b4a Marcin Ciupak 2018-10-17 533 "%s.%d",
> f4952b4a Marcin Ciupak 2018-10-17 534 dev_name(&device->dev),
> f4952b4a Marcin Ciupak 2018-10-17 535 id);
> f4952b4a Marcin Ciupak 2018-10-17 536
> f4952b4a Marcin Ciupak 2018-10-17 537 if (IS_ERR(p->dev)) {
> f4952b4a Marcin Ciupak 2018-10-17 @538 dev_err(&device->dev,
> f4952b4a Marcin Ciupak 2018-10-17 539 "%s: device_create of '%s' failed\n",
> f4952b4a Marcin Ciupak 2018-10-17 540 __func__,
> f4952b4a Marcin Ciupak 2018-10-17 541 dev_name(p->dev));
> f4952b4a Marcin Ciupak 2018-10-17 542 ret = PTR_ERR(p->dev);
> f4952b4a Marcin Ciupak 2018-10-17 543 goto err_ida_remove;
> f4952b4a Marcin Ciupak 2018-10-17 544 }
> f4952b4a Marcin Ciupak 2018-10-17 545
> f4952b4a Marcin Ciupak 2018-10-17 546 cdev_init(&p->cdev, &nrf24_fops);
> f4952b4a Marcin Ciupak 2018-10-17 547 p->cdev.owner = THIS_MODULE;
> f4952b4a Marcin Ciupak 2018-10-17 548 ret = cdev_add(&p->cdev, p->devt, 1);
> f4952b4a Marcin Ciupak 2018-10-17 549 if (ret < 0) {
> f4952b4a Marcin Ciupak 2018-10-17 550 dev_err(&device->dev, "%s: cdev failed\n", __func__);
> f4952b4a Marcin Ciupak 2018-10-17 551 goto err_dev_destroy;
> f4952b4a Marcin Ciupak 2018-10-17 552 }
> f4952b4a Marcin Ciupak 2018-10-17 553
> f4952b4a Marcin Ciupak 2018-10-17 554 dev_dbg(&device->dev,
> f4952b4a Marcin Ciupak 2018-10-17 555 "%s: device created: major(%d), minor(%d)\n",
> f4952b4a Marcin Ciupak 2018-10-17 556 __func__,
> f4952b4a Marcin Ciupak 2018-10-17 557 MAJOR(p->devt),
> f4952b4a Marcin Ciupak 2018-10-17 558 MINOR(p->devt));
> f4952b4a Marcin Ciupak 2018-10-17 559
> f4952b4a Marcin Ciupak 2018-10-17 560 return p;
> f4952b4a Marcin Ciupak 2018-10-17 561
> f4952b4a Marcin Ciupak 2018-10-17 562 err_dev_destroy:
> f4952b4a Marcin Ciupak 2018-10-17 563 device_destroy(nrf24_class, p->devt);
> f4952b4a Marcin Ciupak 2018-10-17 564 err_ida_remove:
> f4952b4a Marcin Ciupak 2018-10-17 565 ida_simple_remove(&nrf24_ida_pipe, MINOR(p->devt));
> f4952b4a Marcin Ciupak 2018-10-17 566 err_free_mem:
> f4952b4a Marcin Ciupak 2018-10-17 567 kfree(p);
> f4952b4a Marcin Ciupak 2018-10-17 568 err_return:
> f4952b4a Marcin Ciupak 2018-10-17 569 return ERR_PTR(ret);
> f4952b4a Marcin Ciupak 2018-10-17 570 }
> f4952b4a Marcin Ciupak 2018-10-17 571
>
> ---
> 0-DAY kernel test infrastructure Open Source Technology Center
> https://lists.01.org/pipermail/kbuild-all Intel Corporation

Interesting, I will take a look on that (and most probably provide a new
patch)

Thanks for the hint Dan!

br,
Marcin