| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Tue, 23 Nov 2021 11:44:32 +0300
From: Dan Carpenter <dan.carpenter@...cle.com>
To: kbuild@...ts.01.org, Andy Lutomirski <luto@...nel.org>
Cc: lkp@...el.com, kbuild-all@...ts.01.org,
linux-kernel@...r.kernel.org
Subject: [luto:sched/lazymm 13/16] kernel/sched/core.c:4982
__change_current_mm() error: uninitialized symbol 'old_active_mm'.
tree: https://git.kernel.org/pub/scm/linux/kernel/git/luto/linux.git sched/lazymm
head: c0d03d4f2778fd0a7c16e69cdfb3f111296129b5
commit: 4863118ffa6fe6af7cfb18c4d4ee6434537b0c8b [13/16] sched, exec: Factor current mm changes out from exec
config: x86_64-randconfig-m001-20211118 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-22) 9.3.0
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@...el.com>
Reported-by: Dan Carpenter <dan.carpenter@...cle.com>
smatch warnings:
kernel/sched/core.c:4982 __change_current_mm() error: uninitialized symbol 'old_active_mm'.
vim +/old_active_mm +4982 kernel/sched/core.c
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4938 void __change_current_mm(struct mm_struct *mm, bool mm_is_brand_new)
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4939 {
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4940 struct task_struct *tsk = current;
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4941 struct mm_struct *old_active_mm, *mm_to_drop = NULL;
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4942
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4943 BUG_ON(!mm); /* likely to cause corruption if we continue */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4944
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4945 /*
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4946 * We do not want to schedule, nor should procfs peek at current->mm
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4947 * while we're modifying it. task_lock() disables preemption and
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4948 * locks against procfs.
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4949 */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4950 task_lock(tsk);
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4951 /*
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4952 * membarrier() requires a full barrier before switching mm.
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4953 */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4954 smp_mb__after_spinlock();
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4955
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4956 local_irq_disable();
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4957
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4958 if (tsk->mm) {
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4959 /* We're detaching from an old mm. Sync stats. */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4960 sync_mm_rss(tsk->mm);
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4961 } else {
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4962 /*
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4963 * Switching from kernel mm to user. Drop the old lazy
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4964 * mm reference.
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4965 */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4966 mm_to_drop = tsk->active_mm;
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4967 }
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4968
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4969 tsk->active_mm = mm;
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4970 WRITE_ONCE(tsk->mm, mm); /* membarrier reads this without locks */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4971 membarrier_update_current_mm(mm);
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4972
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4973 if (mm_is_brand_new) {
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4974 /*
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4975 * For historical reasons, some architectures want IRQs on
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4976 * when activate_mm() is called. If we're going to call
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4977 * activate_mm(), turn on IRQs but leave preemption
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4978 * disabled.
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4979 */
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4980 if (!IS_ENABLED(CONFIG_ARCH_WANT_IRQS_OFF_ACTIVATE_MM))
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4981 local_irq_enable();
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 @4982 activate_mm(old_active_mm, mm);
"old_active_mm" is never initialized.
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4983 if (IS_ENABLED(CONFIG_ARCH_WANT_IRQS_OFF_ACTIVATE_MM))
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4984 local_irq_enable();
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4985 } else {
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4986 switch_mm_irqs_off(old_active_mm, mm, tsk);
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4987 local_irq_enable();
4863118ffa6fe6 kernel/sched/core.c Andy Lutomirski 2021-09-03 4988 }
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
Powered by blists - more mailing lists