Re: [PATCH v4 10/20] nvmem: core: use nvmem_add_one_cell() in nvmem_add_cells_from_of()

[Michael Walle <michael@walle.cc>]

Am 2022-12-03 09:30, schrieb Dan Carpenter:
> Hi Michael,
> 
> https://git-scm.com/docs/git-format-patch#_base_tree_information]
> 
> url:
> https://github.com/intel-lab-lkp/linux/commits/Michael-Walle/nvmem-core-introduce-NVMEM-layouts/20221124-020554
> patch link:
> https://lore.kernel.org/r/20221123180151.2160033-11-michael%40walle.cc
> patch subject: [PATCH v4 10/20] nvmem: core: use nvmem_add_one_cell()
> in nvmem_add_cells_from_of()
> config: i386-randconfig-m021
> compiler: gcc-11 (Debian 11.3.0-8) 11.3.0
> 
> If you fix the issue, kindly add following tag where applicable
> | Reported-by: kernel test robot <lkp@intel.com>
> | Reported-by: Dan Carpenter <error27@gmail.com>
> 
> New smatch warnings:
> drivers/nvmem/core.c:731 nvmem_add_cells_from_of() warn: possible
> memory leak of 'cell'
> 
> Old smatch warnings:
> drivers/nvmem/core.c:735 nvmem_add_cells_from_of() warn: possible
> memory leak of 'cell'
> 
> vim +/cell +731 drivers/nvmem/core.c
> 
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  689  static int
> nvmem_add_cells_from_of(struct nvmem_device *nvmem)
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  690  {
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  691  	struct device
> *dev = &nvmem->dev;
> 7ae6478b304bc0 Srinivas Kandagatla 2021-10-13  692  	struct
> nvmem_cell_entry *cell;
> 18f50dbcfd3676 Michael Walle       2022-11-23  693  	struct device_node 
> *child;
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  694  	const __be32 
> *addr;
> 18f50dbcfd3676 Michael Walle       2022-11-23  695  	int len, ret;
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  696
> 18f50dbcfd3676 Michael Walle       2022-11-23  697
> 	for_each_child_of_node(dev->of_node, child) {
> 18f50dbcfd3676 Michael Walle       2022-11-23  698  		struct
> nvmem_cell_info info = {0};
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  699
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  700  		addr =
> of_get_property(child, "reg", &len);
> 0445efacec75b8 Ahmad Fatoum        2021-01-29  701  		if (!addr)
> 0445efacec75b8 Ahmad Fatoum        2021-01-29  702  			continue;
> 0445efacec75b8 Ahmad Fatoum        2021-01-29  703  		if (len < 2 *
> sizeof(u32)) {
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  704  			dev_err(dev,
> "nvmem: invalid reg on %pOF\n", child);
> 63879e2964bcee Christophe JAILLET  2021-06-11  705  
> 			of_node_put(child);
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  706  			return -EINVAL;
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  707  		}
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  708
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  709  		cell =
> kzalloc(sizeof(*cell), GFP_KERNEL);
> 63879e2964bcee Christophe JAILLET  2021-06-11  710  		if (!cell) {
> 63879e2964bcee Christophe JAILLET  2021-06-11  711  
> 			of_node_put(child);
> e888d445ac33a5 Bartosz Golaszewski 2018-09-21  712  			return -ENOMEM;
> 63879e2964bcee Christophe JAILLET  2021-06-11  713  		}
> 
> Seems like "cell" is not used any more so this just leaks.

Damn, what a stupid bug from me. Thanks for catching this.

-michael