Since commit 07ec77a1d4e8 ("sched: Allow task CPU affinity to be
restricted on asymmetric systems"), the setting and clearing of
user_cpus_ptr are done under pi_lock for arm64 architecture. However,
dup_user_cpus_ptr() accesses user_cpus_ptr without any lock
protection. When racing with the clearing of user_cpus_ptr in
__set_cpus_allowed_ptr_locked(), it can lead to user-after-free and
double-free in arm64 kernel.
Commit 8f9ea86fdf99 ("sched: Always preserve the user requested
cpumask") fixes this problem as user_cpus_ptr, once set, will never
be cleared in a task's lifetime. However, this bug was re-introduced
in commit 851a723e45d1 ("sched: Always clear user_cpus_ptr in
do_set_cpus_allowed()") which allows the clearing of user_cpus_ptr in
do_set_cpus_allowed(). This time, it will affect all arches.
Fix this bug by always clearing the user_cpus_ptr of the newly
cloned/forked task before the copying process starts and check the
user_cpus_ptr state of the source task under pi_lock.
Note to stable, this patch won't be applicable to stable releases.
Just copy the new dup_user_cpus_ptr() function over.
Fixes: 07ec77a1d4e8 ("sched: Allow task CPU affinity to be restricted on asymmetric systems")
Fixes: 851a723e45d1 ("sched: Always clear user_cpus_ptr in do_set_cpus_allowed()")
CC: stable(a)vger.kernel.org
Reported-by: David Wang 王标 <wangbiao3(a)xiaomi.com>
Signed-off-by: Waiman Long <longman(a)redhat.com>
---
kernel/sched/core.c | 34 +++++++++++++++++++++++++++++-----
1 file changed, 29 insertions(+), 5 deletions(-)
[v2: Use data_race() macro as suggested by Will]
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 78b2d5cabcc5..57e5932f81a9 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2612,19 +2612,43 @@ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
int dup_user_cpus_ptr(struct task_struct *dst, struct task_struct *src,
int node)
{
+ cpumask_t *user_mask;
unsigned long flags;
- if (!src->user_cpus_ptr)
+ /*
+ * Always clear dst->user_cpus_ptr first as their user_cpus_ptr's
+ * may differ by now due to racing.
+ */
+ dst->user_cpus_ptr = NULL;
+
+ /*
+ * This check is racy and losing the race is a valid situation.
+ * It is not worth the extra overhead of taking the pi_lock on
+ * every fork/clone.
+ */
+ if (data_race(!src->user_cpus_ptr))
return 0;
- dst->user_cpus_ptr = kmalloc_node(cpumask_size(), GFP_KERNEL, node);
- if (!dst->user_cpus_ptr)
+ user_mask = kmalloc_node(cpumask_size(), GFP_KERNEL, node);
+ if (!user_mask)
return -ENOMEM;
- /* Use pi_lock to protect content of user_cpus_ptr */
+ /*
+ * Use pi_lock to protect content of user_cpus_ptr
+ *
+ * Though unlikely, user_cpus_ptr can be reset to NULL by a concurrent
+ * do_set_cpus_allowed().
+ */
raw_spin_lock_irqsave(&src->pi_lock, flags);
- cpumask_copy(dst->user_cpus_ptr, src->user_cpus_ptr);
+ if (src->user_cpus_ptr) {
+ swap(dst->user_cpus_ptr, user_mask);
+ cpumask_copy(dst->user_cpus_ptr, src->user_cpus_ptr);
+ }
raw_spin_unlock_irqrestore(&src->pi_lock, flags);
+
+ if (unlikely(user_mask))
+ kfree(user_mask);
+
return 0;
}
--
2.31.1
Hi,
It was reported that on 5.15.y there are problems with S3 suspend root
caused to a W6400 problem. [1]
These are fixed by this commit:
a9a1ac44074f ("drm/amd/display: Manually adjust strobe for DCN303")
Can you please bring it to 5.15.y?
Thanks,
[1] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2000299
The memory for "llcc_driv_data" is allocated by the LLCC driver. But when
it is passed as "pvt_info" to the EDAC core, it will get freed during the
qcom_edac driver release. So when the qcom_edac driver gets probed again,
it will try to use the freed data leading to the use-after-free bug.
Fix this by not passing "llcc_driv_data" as pvt_info but rather reference
it using the "platform_data" in the qcom_edac driver.
Cc: <stable(a)vger.kernel.org> # 4.20
Fixes: 27450653f1db ("drivers: edac: Add EDAC driver support for QCOM SoCs")
Reported-by: Steev Klimaszewski <steev(a)kali.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam(a)linaro.org>
---
drivers/edac/qcom_edac.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/edac/qcom_edac.c b/drivers/edac/qcom_edac.c
index 9e77fa84e84f..3256254c3722 100644
--- a/drivers/edac/qcom_edac.c
+++ b/drivers/edac/qcom_edac.c
@@ -252,7 +252,7 @@ dump_syn_reg_values(struct llcc_drv_data *drv, u32 bank, int err_type)
static int
dump_syn_reg(struct edac_device_ctl_info *edev_ctl, int err_type, u32 bank)
{
- struct llcc_drv_data *drv = edev_ctl->pvt_info;
+ struct llcc_drv_data *drv = edev_ctl->dev->platform_data;
int ret;
ret = dump_syn_reg_values(drv, bank, err_type);
@@ -289,7 +289,7 @@ static irqreturn_t
llcc_ecc_irq_handler(int irq, void *edev_ctl)
{
struct edac_device_ctl_info *edac_dev_ctl = edev_ctl;
- struct llcc_drv_data *drv = edac_dev_ctl->pvt_info;
+ struct llcc_drv_data *drv = edac_dev_ctl->dev->platform_data;
irqreturn_t irq_rc = IRQ_NONE;
u32 drp_error, trp_error, i;
int ret;
@@ -358,7 +358,6 @@ static int qcom_llcc_edac_probe(struct platform_device *pdev)
edev_ctl->dev_name = dev_name(dev);
edev_ctl->ctl_name = "llcc";
edev_ctl->panic_on_ue = LLCC_ERP_PANIC_ON_UE;
- edev_ctl->pvt_info = llcc_driv_data;
rc = edac_device_add_device(edev_ctl);
if (rc)
--
2.25.1