Commit 1251201c authored by Ingo Molnar's avatar Ingo Molnar

sched/core: Fix uclamp ABI bug, clean up and robustify sched_read_attr() ABI logic and code

Thadeu Lima de Souza Cascardo reported that 'chrt' broke on recent kernels:

  $ chrt -p $$
  chrt: failed to get pid 26306's policy: Argument list too long

and he has root-caused the bug to the following commit increasing sched_attr
size and breaking sched_read_attr() into returning -EFBIG:

  a509a7cd ("sched/uclamp: Extend sched_setattr() to support utilization clamping")

The other, bigger bug is that the whole sched_getattr() and sched_read_attr()
logic of checking non-zero bits in new ABI components is arguably broken,
and pretty much any extension of the ABI will spuriously break the ABI.
That's way too fragile.

Instead implement the perf syscall's extensible ABI instead, which we
already implement on the sched_setattr() side:

 - if user-attributes have the same size as kernel attributes then the
   logic is unchanged.

 - if user-attributes are larger than the kernel knows about then simply
   skip the extra bits, but set attr->size to the (smaller) kernel size
   so that tooling can (in principle) handle older kernel as well.

 - if user-attributes are smaller than the kernel knows about then just
   copy whatever user-space can accept.

Also clean up the whole logic:

 - Simplify the code flow - there's no need for 'ret' for example.

 - Standardize on 'kattr/uattr' and 'ksize/usize' naming to make sure we
   always know which side we are dealing with.

 - Why is it called 'read' when what it does is to copy to user? This
   code is so far away from VFS read() semantics that the naming is
   actively confusing. Name it sched_attr_copy_to_user() instead, which
   mirrors other copy_to_user() functionality.

 - Move the attr->size assignment from the head of sched_getattr() to the
   sched_attr_copy_to_user() function. Nothing else within the kernel
   should care about the size of the structure.

With these fixes the sched_getattr() syscall now nicely supports an
extensible ABI in both a forward and backward compatible fashion, and
will also fix the chrt bug.

As an added bonus the bogus -EFBIG return is removed as well, which as
Thadeu noted should have been -E2BIG to begin with.
Reported-by: default avatarThadeu Lima de Souza Cascardo <cascardo@canonical.com>
Tested-by: default avatarDietmar Eggemann <dietmar.eggemann@arm.com>
Tested-by: default avatarThadeu Lima de Souza Cascardo <cascardo@canonical.com>
Acked-by: default avatarThadeu Lima de Souza Cascardo <cascardo@canonical.com>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Patrick Bellasi <patrick.bellasi@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: a509a7cd ("sched/uclamp: Extend sched_setattr() to support utilization clamping")
Link: https://lkml.kernel.org/r/20190904075532.GA26751@gmail.comSigned-off-by: default avatarIngo Molnar <mingo@kernel.org>
parent 5e2d2cc2
...@@ -5105,37 +5105,40 @@ SYSCALL_DEFINE2(sched_getparam, pid_t, pid, struct sched_param __user *, param) ...@@ -5105,37 +5105,40 @@ SYSCALL_DEFINE2(sched_getparam, pid_t, pid, struct sched_param __user *, param)
return retval; return retval;
} }
static int sched_read_attr(struct sched_attr __user *uattr, /*
struct sched_attr *attr, * Copy the kernel size attribute structure (which might be larger
* than what user-space knows about) to user-space.
*
* Note that all cases are valid: user-space buffer can be larger or
* smaller than the kernel-space buffer. The usual case is that both
* have the same size.
*/
static int
sched_attr_copy_to_user(struct sched_attr __user *uattr,
struct sched_attr *kattr,
unsigned int usize) unsigned int usize)
{ {
int ret; unsigned int ksize = sizeof(*kattr);
if (!access_ok(uattr, usize)) if (!access_ok(uattr, usize))
return -EFAULT; return -EFAULT;
/* /*
* If we're handed a smaller struct than we know of, * sched_getattr() ABI forwards and backwards compatibility:
* ensure all the unknown bits are 0 - i.e. old *
* user-space does not get uncomplete information. * If usize == ksize then we just copy everything to user-space and all is good.
*
* If usize < ksize then we only copy as much as user-space has space for,
* this keeps ABI compatibility as well. We skip the rest.
*
* If usize > ksize then user-space is using a newer version of the ABI,
* which part the kernel doesn't know about. Just ignore it - tooling can
* detect the kernel's knowledge of attributes from the attr->size value
* which is set to ksize in this case.
*/ */
if (usize < sizeof(*attr)) { kattr->size = min(usize, ksize);
unsigned char *addr;
unsigned char *end;
addr = (void *)attr + usize; if (copy_to_user(uattr, kattr, kattr->size))
end = (void *)attr + sizeof(*attr);
for (; addr < end; addr++) {
if (*addr)
return -EFBIG;
}
attr->size = usize;
}
ret = copy_to_user(uattr, attr, attr->size);
if (ret)
return -EFAULT; return -EFAULT;
return 0; return 0;
...@@ -5145,20 +5148,18 @@ static int sched_read_attr(struct sched_attr __user *uattr, ...@@ -5145,20 +5148,18 @@ static int sched_read_attr(struct sched_attr __user *uattr,
* sys_sched_getattr - similar to sched_getparam, but with sched_attr * sys_sched_getattr - similar to sched_getparam, but with sched_attr
* @pid: the pid in question. * @pid: the pid in question.
* @uattr: structure containing the extended parameters. * @uattr: structure containing the extended parameters.
* @size: sizeof(attr) for fwd/bwd comp. * @usize: sizeof(attr) that user-space knows about, for forwards and backwards compatibility.
* @flags: for future extension. * @flags: for future extension.
*/ */
SYSCALL_DEFINE4(sched_getattr, pid_t, pid, struct sched_attr __user *, uattr, SYSCALL_DEFINE4(sched_getattr, pid_t, pid, struct sched_attr __user *, uattr,
unsigned int, size, unsigned int, flags) unsigned int, usize, unsigned int, flags)
{ {
struct sched_attr attr = { struct sched_attr kattr = { };
.size = sizeof(struct sched_attr),
};
struct task_struct *p; struct task_struct *p;
int retval; int retval;
if (!uattr || pid < 0 || size > PAGE_SIZE || if (!uattr || pid < 0 || usize > PAGE_SIZE ||
size < SCHED_ATTR_SIZE_VER0 || flags) usize < SCHED_ATTR_SIZE_VER0 || flags)
return -EINVAL; return -EINVAL;
rcu_read_lock(); rcu_read_lock();
...@@ -5171,25 +5172,24 @@ SYSCALL_DEFINE4(sched_getattr, pid_t, pid, struct sched_attr __user *, uattr, ...@@ -5171,25 +5172,24 @@ SYSCALL_DEFINE4(sched_getattr, pid_t, pid, struct sched_attr __user *, uattr,
if (retval) if (retval)
goto out_unlock; goto out_unlock;
attr.sched_policy = p->policy; kattr.sched_policy = p->policy;
if (p->sched_reset_on_fork) if (p->sched_reset_on_fork)
attr.sched_flags |= SCHED_FLAG_RESET_ON_FORK; kattr.sched_flags |= SCHED_FLAG_RESET_ON_FORK;
if (task_has_dl_policy(p)) if (task_has_dl_policy(p))
__getparam_dl(p, &attr); __getparam_dl(p, &kattr);
else if (task_has_rt_policy(p)) else if (task_has_rt_policy(p))
attr.sched_priority = p->rt_priority; kattr.sched_priority = p->rt_priority;
else else
attr.sched_nice = task_nice(p); kattr.sched_nice = task_nice(p);
#ifdef CONFIG_UCLAMP_TASK #ifdef CONFIG_UCLAMP_TASK
attr.sched_util_min = p->uclamp_req[UCLAMP_MIN].value; kattr.sched_util_min = p->uclamp_req[UCLAMP_MIN].value;
attr.sched_util_max = p->uclamp_req[UCLAMP_MAX].value; kattr.sched_util_max = p->uclamp_req[UCLAMP_MAX].value;
#endif #endif
rcu_read_unlock(); rcu_read_unlock();
retval = sched_read_attr(uattr, &attr, size); return sched_attr_copy_to_user(uattr, &kattr, usize);
return retval;
out_unlock: out_unlock:
rcu_read_unlock(); rcu_read_unlock();
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment