| Message ID | 20231113200742.3593548-3-harshit.m.mogalapalli@oracle.com (mailing list archive) |
|---|---|
| State | Accepted, archived |
| Delegated to: | Ilpo Järvinen |
| Headers | show |
| Series | [v4,1/4] platform/x86: hp-bioscfg: Simplify return check in hp_add_other_attributes() | expand |
On Mon, 13 Nov 2023, Harshit Mogalapalli wrote: > 'attr_name_kobj' is allocated using kzalloc, but on all the error paths > it is not freed, hence we have a memory leak. > > Fix the error path before kobject_init_and_add() by adding kfree(). > > kobject_put() must be always called after passing the object to > kobject_init_and_add(). Only the error path which is immediately next > to kobject_init_and_add() calls kobject_put() and not any other error > path after it. > > Fix the error handling after kobject_init_and_add() by moving the > kobject_put() into the goto label err_other_attr_init that is already > used by all the error paths after kobject_init_and_add(). > > Fixes: a34fc329b189 ("platform/x86: hp-bioscfg: bioscfg") > Reported-by: kernel test robot <lkp@intel.com> > Reported-by: Dan Carpenter <error27@gmail.com> > Closes: https://lore.kernel.org/r/202309201412.on0VXJGo-lkp@intel.com/ > Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com> > --- > This is based on static analysis, only compile tested. > > v3->v4: Add more explicit statement on how we are fixing it, suggested > by Ilpo Thanks a lot, this looks fine too now. Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Hi Ilpo, On 11/14/23 11:31, Ilpo Järvinen wrote: > On Mon, 13 Nov 2023, Harshit Mogalapalli wrote: > >> 'attr_name_kobj' is allocated using kzalloc, but on all the error paths >> it is not freed, hence we have a memory leak. >> >> Fix the error path before kobject_init_and_add() by adding kfree(). >> >> kobject_put() must be always called after passing the object to >> kobject_init_and_add(). Only the error path which is immediately next >> to kobject_init_and_add() calls kobject_put() and not any other error >> path after it. >> >> Fix the error handling after kobject_init_and_add() by moving the >> kobject_put() into the goto label err_other_attr_init that is already >> used by all the error paths after kobject_init_and_add(). >> >> Fixes: a34fc329b189 ("platform/x86: hp-bioscfg: bioscfg") >> Reported-by: kernel test robot <lkp@intel.com> >> Reported-by: Dan Carpenter <error27@gmail.com> >> Closes: https://lore.kernel.org/r/202309201412.on0VXJGo-lkp@intel.com/ >> Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com> >> --- >> This is based on static analysis, only compile tested. >> >> v3->v4: Add more explicit statement on how we are fixing it, suggested >> by Ilpo > > Thanks a lot, this looks fine too now. > > Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Thank you for reviewing this series. I believe that it is best to submit this as fixes for the current cycle. Under the assumption that you agree with this I've delegated these 4 patches to you (Ilpo) in patchwork. Regards, Hans
On Tue, 14 Nov 2023, Hans de Goede wrote: > On 11/14/23 11:31, Ilpo Järvinen wrote: > > On Mon, 13 Nov 2023, Harshit Mogalapalli wrote: > > > >> 'attr_name_kobj' is allocated using kzalloc, but on all the error paths > >> it is not freed, hence we have a memory leak. > >> > >> Fix the error path before kobject_init_and_add() by adding kfree(). > >> > >> kobject_put() must be always called after passing the object to > >> kobject_init_and_add(). Only the error path which is immediately next > >> to kobject_init_and_add() calls kobject_put() and not any other error > >> path after it. > >> > >> Fix the error handling after kobject_init_and_add() by moving the > >> kobject_put() into the goto label err_other_attr_init that is already > >> used by all the error paths after kobject_init_and_add(). > >> > >> Fixes: a34fc329b189 ("platform/x86: hp-bioscfg: bioscfg") > >> Reported-by: kernel test robot <lkp@intel.com> > >> Reported-by: Dan Carpenter <error27@gmail.com> > >> Closes: https://lore.kernel.org/r/202309201412.on0VXJGo-lkp@intel.com/ > >> Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com> > >> --- > >> This is based on static analysis, only compile tested. > >> > >> v3->v4: Add more explicit statement on how we are fixing it, suggested > >> by Ilpo > > > > Thanks a lot, this looks fine too now. > > > > Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> > > Thank you for reviewing this series. I believe that it is best > to submit this as fixes for the current cycle. > > Under the assumption that you agree with this I've delegated > these 4 patches to you (Ilpo) in patchwork. Yes, I'll take care of it later this week.
diff --git a/drivers/platform/x86/hp/hp-bioscfg/bioscfg.c b/drivers/platform/x86/hp/hp-bioscfg/bioscfg.c index a3599498c4e8..6ddca857cc4d 100644 --- a/drivers/platform/x86/hp/hp-bioscfg/bioscfg.c +++ b/drivers/platform/x86/hp/hp-bioscfg/bioscfg.c @@ -575,77 +575,79 @@ static void release_attributes_data(void) /** * hp_add_other_attributes() - Initialize HP custom attributes not * reported by BIOS and required to support Secure Platform and Sure * Start. * * @attr_type: Custom HP attribute not reported by BIOS * * Initialize all 2 types of attributes: Platform and Sure Start * object. Populates each attribute types respective properties * under sysfs files. * * Returns zero(0) if successful. Otherwise, a negative value. */ static int hp_add_other_attributes(int attr_type) { struct kobject *attr_name_kobj; union acpi_object *obj = NULL; int ret; char *attr_name; attr_name_kobj = kzalloc(sizeof(*attr_name_kobj), GFP_KERNEL); if (!attr_name_kobj) return -ENOMEM; mutex_lock(&bioscfg_drv.mutex); /* Check if attribute type is supported */ switch (attr_type) { case HPWMI_SECURE_PLATFORM_TYPE: attr_name_kobj->kset = bioscfg_drv.authentication_dir_kset; attr_name = SPM_STR; break; case HPWMI_SURE_START_TYPE: attr_name_kobj->kset = bioscfg_drv.main_dir_kset; attr_name = SURE_START_STR; break; default: pr_err("Error: Unknown attr_type: %d\n", attr_type); ret = -EINVAL; - goto err_other_attr_init; + kfree(attr_name_kobj); + goto unlock_drv_mutex; } ret = kobject_init_and_add(attr_name_kobj, &attr_name_ktype, NULL, "%s", attr_name); if (ret) { pr_err("Error encountered [%d]\n", ret); - kobject_put(attr_name_kobj); goto err_other_attr_init; } /* Populate attribute data */ switch (attr_type) { case HPWMI_SECURE_PLATFORM_TYPE: ret = hp_populate_secure_platform_data(attr_name_kobj); break; case HPWMI_SURE_START_TYPE: ret = hp_populate_sure_start_data(attr_name_kobj); break; default: ret = -EINVAL; } if (ret) goto err_other_attr_init; mutex_unlock(&bioscfg_drv.mutex); return 0; err_other_attr_init: + kobject_put(attr_name_kobj); +unlock_drv_mutex: mutex_unlock(&bioscfg_drv.mutex); kfree(obj); return ret; }
'attr_name_kobj' is allocated using kzalloc, but on all the error paths it is not freed, hence we have a memory leak. Fix the error path before kobject_init_and_add() by adding kfree(). kobject_put() must be always called after passing the object to kobject_init_and_add(). Only the error path which is immediately next to kobject_init_and_add() calls kobject_put() and not any other error path after it. Fix the error handling after kobject_init_and_add() by moving the kobject_put() into the goto label err_other_attr_init that is already used by all the error paths after kobject_init_and_add(). Fixes: a34fc329b189 ("platform/x86: hp-bioscfg: bioscfg") Reported-by: kernel test robot <lkp@intel.com> Reported-by: Dan Carpenter <error27@gmail.com> Closes: https://lore.kernel.org/r/202309201412.on0VXJGo-lkp@intel.com/ Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com> --- This is based on static analysis, only compile tested. v3->v4: Add more explicit statement on how we are fixing it, suggested by Ilpo --- drivers/platform/x86/hp/hp-bioscfg/bioscfg.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)