On 2019-09-05, Peter Zijlstra [off-list ref] wrote:

On Thu, Sep 05, 2019 at 07:26:22PM +1000, Aleksa Sarai wrote:

quoted

On 2019-09-05, Peter Zijlstra [off-list ref] wrote:

quoted

On Thu, Sep 05, 2019 at 06:19:22AM +1000, Aleksa Sarai wrote:

quoted

+/**
+ * copy_struct_to_user: copy a struct to user space
+ * @dst:   Destination address, in user space.
+ * @usize: Size of @dst struct.
+ * @src:   Source address, in kernel space.
+ * @ksize: Size of @src struct.
+ *
+ * Copies a struct from kernel space to user space, in a way that guarantees
+ * backwards-compatibility for struct syscall arguments (as long as future
+ * struct extensions are made such that all new fields are *appended* to the
+ * old struct, and zeroed-out new fields have the same meaning as the old
+ * struct).
+ *
+ * @ksize is just sizeof(*dst), and @usize should've been passed by user space.
+ * The recommended usage is something like the following:
+ *
+ *   SYSCALL_DEFINE2(foobar, struct foo __user *, uarg, size_t, usize)
+ *   {
+ *      int err;
+ *      struct foo karg = {};
+ *
+ *      // do something with karg
+ *
+ *      err = copy_struct_to_user(uarg, usize, &karg, sizeof(karg));
+ *      if (err)
+ *        return err;
+ *
+ *      // ...
+ *   }
+ *
+ * There are three cases to consider:
+ *  * If @usize == @ksize, then it's copied verbatim.
+ *  * If @usize < @ksize, then kernel space is "returning" a newer struct to an
+ *    older user space. In order to avoid user space getting incomplete
+ *    information (new fields might be important), all trailing bytes in @src
+ *    (@ksize - @usize) must be zerored

s/zerored/zero/, right?

It should've been "zeroed".

That reads wrong to me; that way it reads like this function must take
that action and zero out the 'rest'; which is just wrong.

This function must verify those bytes are zero, not make them zero.

Right, in my head I was thinking "must have been zeroed" which isn't
what it says. I'll switch to "zero".

quoted

quoted

quoted

                                         , otherwise -EFBIG is returned.

'Funny' that, copy_struct_from_user() below seems to use E2BIG.

This is a copy of the semantics that sched_[sg]etattr(2) uses -- E2BIG for
a "too big" struct passed to the kernel, and EFBIG for a "too big"
struct passed to user-space. I would personally have preferred EMSGSIZE
instead of EFBIG, but felt using the existing error codes would be less
confusing.

Sadly a recent commit:

  1251201c0d34 ("sched/core: Fix uclamp ABI bug, clean up and robustify sched_read_attr() ABI logic and code")

Made the situation even 'worse'.

I hadn't seen this patch before, and I have a few questions taking a
look at it:

 * An error code for a particular behaviour was changed (EFBIG ->
   E2BIG). Is this not a userspace breakage (I know Linus went ballistic
   about something similar a while ago[1]), or did I misunderstand what
   the issue was in [1]?
   * At the risk of bike-shedding -- of we are changing it, wouldn't
	 -EMSGSIZE be more appropriate? To be fair, picking errno values has
	 always been more of an art than a science, but to my ears "Argument
	 list too long" doesn't make too much sense in the context of
	 "returning" a struct back to userspace (and the cause of the error
	 is that the argument passed by user space *isn't big enough*). If
	 there was an E2SMALL that would also work. ;)

 * Do you want me to write a patch based on that, to switch it to
   copy_struct_to_user()?

 * That patch removes the "are there non-zero bytes in the tail that
   userspace won't know about" check (which I have included in mine). I
   understand that this caused issues specifically with sched_getattr(2)
   due to the default value not being zero -- how should we rectify that
   (given that we'd hopefully want to port everyone who uses that
   interface to copy_struct_{to,from}_user())?

 * Given that the [uk]attr->size construct is pretty important to the
   usability of the sched and perf interfaces, should we require (or
   encourage) it for all struct-extension syscall setups?

quoted

quoted

quoted

+	if (unlikely(!access_ok(src, usize)))
+		return -EFAULT;
+
+	/* Deal with trailing bytes. */
+	if (usize < ksize)
+		memset(dst + size, 0, rest);
+	else if (usize > ksize) {
+		const void __user *addr = src + size;
+		char buffer[BUFFER_SIZE] = {};

Isn't that too big for on-stack?

Is a 64-byte buffer too big? I picked the number "at random" to be the
size of a cache line, but I could shrink it down to 32 bytes if the size
is an issue (I wanted to avoid needless allocations -- hence it being
on-stack).

Ah, my ctags gave me a definition of BUFFER_SIZE that was 512. I suppose
64 should be OK.

Good to know, though I'll rename it to avoid confusion.

[1]: https://lore.kernel.org/lkml/CA+55aFy98A+LJK4+GWMcbzaa1zsPBRo76q+ioEjbx-uaMKH6Uw@mail.gmail.com/ (local)

-- 
Aleksa Sarai
Senior Software Engineer (Containers)
SUSE Linux GmbH
<https://www.cyphar.com/>