[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <155085b0-d2e4-4503-b75b-1a71193c2587@kili.mountain>
Date: Tue, 21 Mar 2023 07:53:32 +0300
From: Dan Carpenter <error27@...il.com>
To: oe-kbuild@...ts.linux.dev, Rob Clark <robdclark@...il.com>,
dri-devel@...ts.freedesktop.org
Cc: lkp@...el.com, oe-kbuild-all@...ts.linux.dev,
Rob Clark <robdclark@...omium.org>,
Len Brown <len.brown@...el.com>, Pavel Machek <pavel@....cz>,
"Rafael J. Wysocki" <rafael@...nel.org>,
linux-arm-msm@...r.kernel.org,
"open list:POWER MANAGEMENT CORE" <linux-pm@...r.kernel.org>,
open list <linux-kernel@...r.kernel.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
freedreno@...ts.freedesktop.org
Subject: Re: [PATCH v2 18/23] PM / QoS: Decouple request alloc from
dev_pm_qos_mtx
Hi Rob,
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Rob-Clark/drm-msm-Pre-allocate-hw_fence/20230320-224826
base: git://anongit.freedesktop.org/drm/drm-misc drm-misc-next
patch link: https://lore.kernel.org/r/20230320144356.803762-19-robdclark%40gmail.com
patch subject: [PATCH v2 18/23] PM / QoS: Decouple request alloc from dev_pm_qos_mtx
config: arm64-randconfig-m041-20230319 (https://download.01.org/0day-ci/archive/20230321/202303211207.mUCSt3CK-lkp@intel.com/config)
compiler: aarch64-linux-gcc (GCC) 12.1.0
If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@...el.com>
| Reported-by: Dan Carpenter <error27@...il.com>
| Link: https://lore.kernel.org/r/202303211207.mUCSt3CK-lkp@intel.com/
smatch warnings:
drivers/base/power/qos.c:947 dev_pm_qos_update_user_latency_tolerance() error: uninitialized symbol 'req'.
drivers/base/power/qos.c:975 dev_pm_qos_update_user_latency_tolerance() warn: possible memory leak of 'req'
vim +/req +947 drivers/base/power/qos.c
2d984ad132a87c Rafael J. Wysocki 2014-02-11 923 int dev_pm_qos_update_user_latency_tolerance(struct device *dev, s32 val)
2d984ad132a87c Rafael J. Wysocki 2014-02-11 924 {
2d7e4629d7265d Rob Clark 2023-03-20 925 struct dev_pm_qos_request *req = NULL;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 926 int ret;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 927
00dd582e52a535 Rob Clark 2023-03-20 928 ret = dev_pm_qos_constraints_ensure_allocated(dev);
00dd582e52a535 Rob Clark 2023-03-20 929 if (ret)
00dd582e52a535 Rob Clark 2023-03-20 930 return ret;
00dd582e52a535 Rob Clark 2023-03-20 931
2d7e4629d7265d Rob Clark 2023-03-20 932 if (!dev->power.qos->latency_tolerance_req)
2d7e4629d7265d Rob Clark 2023-03-20 933 req = kzalloc(sizeof(*req), GFP_KERNEL);
2d7e4629d7265d Rob Clark 2023-03-20 934
2d984ad132a87c Rafael J. Wysocki 2014-02-11 935 mutex_lock(&dev_pm_qos_mtx);
2d984ad132a87c Rafael J. Wysocki 2014-02-11 936
00dd582e52a535 Rob Clark 2023-03-20 937 if (!dev->power.qos->latency_tolerance_req) {
2d984ad132a87c Rafael J. Wysocki 2014-02-11 938 struct dev_pm_qos_request *req;
This "req" shadows the ealier req.
2d984ad132a87c Rafael J. Wysocki 2014-02-11 939
2d984ad132a87c Rafael J. Wysocki 2014-02-11 940 if (val < 0) {
80a6f7c79b7822 Andrew Lutomirski 2016-11-29 941 if (val == PM_QOS_LATENCY_TOLERANCE_NO_CONSTRAINT)
80a6f7c79b7822 Andrew Lutomirski 2016-11-29 942 ret = 0;
80a6f7c79b7822 Andrew Lutomirski 2016-11-29 943 else
2d984ad132a87c Rafael J. Wysocki 2014-02-11 944 ret = -EINVAL;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 945 goto out;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 946 }
2d984ad132a87c Rafael J. Wysocki 2014-02-11 @947 if (!req) {
So it leads to an unintialized variable and a leak.
2d984ad132a87c Rafael J. Wysocki 2014-02-11 948 ret = -ENOMEM;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 949 goto out;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 950 }
2d984ad132a87c Rafael J. Wysocki 2014-02-11 951 ret = __dev_pm_qos_add_request(dev, req, DEV_PM_QOS_LATENCY_TOLERANCE, val);
2d984ad132a87c Rafael J. Wysocki 2014-02-11 952 if (ret < 0) {
2d984ad132a87c Rafael J. Wysocki 2014-02-11 953 kfree(req);
2d984ad132a87c Rafael J. Wysocki 2014-02-11 954 goto out;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 955 }
2d984ad132a87c Rafael J. Wysocki 2014-02-11 956 dev->power.qos->latency_tolerance_req = req;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 957 } else {
2d7e4629d7265d Rob Clark 2023-03-20 958 /*
2d7e4629d7265d Rob Clark 2023-03-20 959 * If we raced with another thread to allocate the request,
2d7e4629d7265d Rob Clark 2023-03-20 960 * simply free the redundant allocation and move on.
2d7e4629d7265d Rob Clark 2023-03-20 961 */
2d7e4629d7265d Rob Clark 2023-03-20 962 if (req)
2d7e4629d7265d Rob Clark 2023-03-20 963 kfree(req);
2d7e4629d7265d Rob Clark 2023-03-20 964
2d984ad132a87c Rafael J. Wysocki 2014-02-11 965 if (val < 0) {
2d984ad132a87c Rafael J. Wysocki 2014-02-11 966 __dev_pm_qos_drop_user_request(dev, DEV_PM_QOS_LATENCY_TOLERANCE);
2d984ad132a87c Rafael J. Wysocki 2014-02-11 967 ret = 0;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 968 } else {
2d984ad132a87c Rafael J. Wysocki 2014-02-11 969 ret = __dev_pm_qos_update_request(dev->power.qos->latency_tolerance_req, val);
2d984ad132a87c Rafael J. Wysocki 2014-02-11 970 }
2d984ad132a87c Rafael J. Wysocki 2014-02-11 971 }
2d984ad132a87c Rafael J. Wysocki 2014-02-11 972
2d984ad132a87c Rafael J. Wysocki 2014-02-11 973 out:
2d984ad132a87c Rafael J. Wysocki 2014-02-11 974 mutex_unlock(&dev_pm_qos_mtx);
2d984ad132a87c Rafael J. Wysocki 2014-02-11 @975 return ret;
2d984ad132a87c Rafael J. Wysocki 2014-02-11 976 }
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests
Powered by blists - more mailing lists