dma-buf: dma-heap: close installed fd when copy_to_user() fails

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, the ioctl returns -EFAULT and userspace never learns the fd number, but the fd (and the underlying dma-buf reference) remain in the caller's fd table and are leaked for the lifetime of the process.

The failure is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Each such ioctl leaks one dmabuf fd; repeating the call quickly fills /proc/<pid>/fd with anonymous "/dmabuf:" entries that only go away when the process exits.

Fix it by closing the installed fd (and clearing the fd field of the kernel-side copy) when copy_to_user() fails after a successful allocation, so the error path matches what userspace observes: no fd was returned, therefore no fd is left behind.

---

Reproducer (full source, gcc -o poc poc.c; run as root):

    // poc.c -- leak one dma-buf fd per DMA_HEAP_IOCTL_ALLOC
    //          when copy_to_user() fails
    #include <fcntl.h>
    #include <stdio.h>
    #include <stdlib.h>
    #include <string.h>
    #include <unistd.h>
    #include <sys/ioctl.h>
    #include <sys/mman.h>
    #include <linux/dma-heap.h>

    int main(int argc, char **argv)
    {
        int n = argc > 1 ? atoi(argv[1]) : 100;
        long ps = sysconf(_SC_PAGESIZE);

        int heap = open("/dev/dma_heap/system", O_RDWR | O_CLOEXEC);
        if (heap < 0)
            return perror("open"), 1;

        for (int i = 0; i < n; i++) {
            /* Put a valid request in a page, then make the page
             * read-only: copy_from_user() still succeeds and the
             * dma-buf is allocated and fd_install()'d, but the
             * trailing copy_to_user() fails and the fd, never
             * returned to us, is leaked.
             */
            struct dma_heap_allocation_data *req =
                mmap(NULL, ps, PROT_READ | PROT_WRITE,
                     MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);

            memset(req, 0, sizeof(*req));
            req->len = ps;
            req->fd_flags = O_RDWR | O_CLOEXEC;

            mprotect(req, ps, PROT_READ);
            ioctl(heap, DMA_HEAP_IOCTL_ALLOC, req);  /* -EFAULT */
            munmap(req, ps);
        }

        printf("done: check ls -l /proc/%d/fd for %d leaked fds\n",
               getpid(), n);
        pause();
        return 0;
    }

Before the fix, ./poc 10 leaves 10 anonymous dmabuf fds in the caller's fd table:

    # ls -l /proc/$(pgrep poc)/fd
    lrwx------ 1 root root 64 Jan  1 00:03 3 -> /dev/dma_heap/system
    lrwx------ 1 root root 64 Jan  1 00:03 4 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 5 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 6 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 7 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 8 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 9 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 10 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 11 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 12 -> /dmabuf:
    lrwx------ 1 root root 64 Jan  1 00:03 13 -> /dmabuf:

After the fix, only /dev/dma_heap/system remains open; the anonymous "/dmabuf:" entries are gone.

 drivers/dma-buf/dma-heap.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

index a76bf3f8b071..0dd7a84b06bf 100644
@@ -18,6 +18,7 @@
 #include <linux/uaccess.h>
 #include <linux/xarray.h>
 #include <uapi/linux/dma-heap.h>
+#include <linux/fdtable.h>
 
 #define DEVNAME "dma_heap"
 

@@ -181,8 +182,16 @@ static long dma_heap_ioctl(struct file *file, unsigned int ucmd,
 		goto err;
 	}
 
-	if (copy_to_user((void __user *)arg, kdata, out_size) != 0)
+	if (copy_to_user((void __user *)arg, kdata, out_size) != 0) {
+		if (kcmd == DMA_HEAP_IOCTL_ALLOC && ret == 0) {
+			struct dma_heap_allocation_data *h = (void *)kdata;
+
+			close_fd(h->fd);
+			h->fd = -1;
+		}
 		ret = -EFAULT;
+	}
+
 err:
 	if (kdata != stack_kdata)
 		kfree(kdata);

2.34.1

This came up many many times before and is the completely wrong approach to handle that.

Background is that this opens up a race window between install_fd() and close_fd() which is much more worse than the original issue and can be abused for a couple of different things.

If I'm not completely mistaken we now have automated checkers in place which find such use case in the upstream kernel and complain about it.

IIRC the correct approach is to call get_unused_fd_flags() to get an fd, do your things, copy it to user space etc... and then finally do the fd_install(fd, dmabuf->file) when nothing can go wrong any more.

If an error happens use put_unused_fd() to clean up.

Regards,
Christian.

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. fd_install() publishes the fd -- from here on the
ioctl cannot fail.

To make this possible, dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4. The fd is written into the kdata buffer before copy_to_user() so the reserved fd number reaches userspace atomically with the install.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 1 file changed, 40 insertions(+), 40 deletions(-)

index a76bf3f8b071..0a9bf62eb06c 100644
@@ -55,33 +55,6 @@ MODULE_PARM_DESC(mem_accounting,
 		 "Enable cgroup-based memory accounting for dma-buf heap allocations (default=false).");
 EXPORT_SYMBOL_NS_GPL(mem_accounting, "DMA_BUF_HEAP");
 
-static int dma_heap_buffer_alloc(struct dma_heap *heap, size_t len,
-				 u32 fd_flags,
-				 u64 heap_flags)
-{
-	struct dma_buf *dmabuf;
-	int fd;
-
-	/*
-	 * Allocations from all heaps have to begin
-	 * and end on page boundaries.
-	 */
-	len = PAGE_ALIGN(len);
-	if (!len)
-		return -EINVAL;
-
-	dmabuf = heap->ops->allocate(heap, len, fd_flags, heap_flags);
-	if (IS_ERR(dmabuf))
-		return PTR_ERR(dmabuf);
-
-	fd = dma_buf_fd(dmabuf, fd_flags);
-	if (fd < 0) {
-		dma_buf_put(dmabuf);
-		/* just return, as put will call release and that will free */
-	}
-	return fd;
-}
-
 static int dma_heap_open(struct inode *inode, struct file *file)
 {
 	struct dma_heap *heap;

@@ -99,30 +72,42 @@ static int dma_heap_open(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static long dma_heap_ioctl_allocate(struct file *file, void *data)
+static struct dma_buf *dma_heap_ioctl_allocate(struct file *file, void *data)
 {
 	struct dma_heap_allocation_data *heap_allocation = data;
 	struct dma_heap *heap = file->private_data;
+	struct dma_buf *dmabuf;
 	int fd;
+	size_t len;
 
 	if (heap_allocation->fd)
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
 
 	if (heap_allocation->fd_flags & ~DMA_HEAP_VALID_FD_FLAGS)
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
 
 	if (heap_allocation->heap_flags & ~DMA_HEAP_VALID_HEAP_FLAGS)
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
+
+	len = PAGE_ALIGN(heap_allocation->len);
+	if (!len)
+		return ERR_PTR(-EINVAL);
+
+	dmabuf = heap->ops->allocate(heap, len, heap_allocation->fd_flags,
+				     heap_allocation->heap_flags);
 
-	fd = dma_heap_buffer_alloc(heap, heap_allocation->len,
-				   heap_allocation->fd_flags,
-				   heap_allocation->heap_flags);
-	if (fd < 0)
-		return fd;
+	if (IS_ERR(dmabuf))
+		return dmabuf;
+
+	fd = get_unused_fd_flags(heap_allocation->fd_flags);
+	if (fd < 0) {
+		dma_buf_put(dmabuf);
+		return ERR_PTR(fd);
+	}
 
 	heap_allocation->fd = fd;
 
-	return 0;
+	return dmabuf;
 }
 
 static unsigned int dma_heap_ioctl_cmds[] = {

@@ -138,6 +123,8 @@ static long dma_heap_ioctl(struct file *file, unsigned int ucmd,
 	unsigned int in_size, out_size, drv_size, ksize;
 	int nr = _IOC_NR(ucmd);
 	int ret = 0;
+	int fd;
+	struct dma_buf *dmabuf;
 
 	if (nr >= ARRAY_SIZE(dma_heap_ioctl_cmds))
 		return -EINVAL;

@@ -174,15 +161,28 @@ static long dma_heap_ioctl(struct file *file, unsigned int ucmd,
 
 	switch (kcmd) {
 	case DMA_HEAP_IOCTL_ALLOC:
-		ret = dma_heap_ioctl_allocate(file, kdata);
+		dmabuf = dma_heap_ioctl_allocate(file, kdata);
+
+		if (IS_ERR(dmabuf)) {
+			ret = PTR_ERR(dmabuf);
+			break;
+		}
+
+		fd = ((struct dma_heap_allocation_data *)kdata)->fd;
+		if (copy_to_user((void __user *)arg, kdata, out_size) != 0) {
+			put_unused_fd(fd);
+			dma_buf_put(dmabuf);
+			ret = -EFAULT;
+		} else {
+			fd_install(fd, dmabuf->file);
+		}
+
 		break;
 	default:
 		ret = -ENOTTY;
 		goto err;
 	}
 
-	if (copy_to_user((void __user *)arg, kdata, out_size) != 0)
-		ret = -EFAULT;
 err:
 	if (kdata != stack_kdata)
 		kfree(kdata);

2.34.1

Hi Xiang Gao,

It looks like this change in its current form breaks the export tracing that you introduced. It looks fixable by adding a call to DMA_BUF_TRACE in after copy_to_user in dma_heap_ioctl. However I'm wondering if you're aware of the dmabuf iterator that traverses the dmabuf_list? I've backported it to 5.10, so it's available on all our
kernels: https://r.android.com/q/Iaeda328f7693acc1506269c12cac5e7955e5c091
Someone else from Xiaomi even helped with the 6.12 backport. So I guess I'm wondering if the dmabuf iterator would work for you instead of the tracepoints.

Thanks,
T.J.

Hi T.J.,

Thanks for the review.

You are right, the export tracing was inadvertently lost when splitting dma_buf_fd() apart. I've fixed it in v3 by introducing a new dma_buf_fd_install() helper in dma-buf.c that wraps the DMA_BUF_TRACE call together with fd_install(), and using that in dma-heap.c instead of a bare fd_install(). The trace event fires after copy_to_user() succeeds, just before the fd becomes visible to userspace — the same observable moment as before.

Baineng

Thanks Baineng, that LGTM. Please have my RB:

You'll probably want to resend this with a "[PATCH v3]" subject though.

Originally I wasn't sure if we really needed dma_buf_fd_install exported, but I looked around and see that fastrpc_dmabuf_alloc (drivers/misc/fastrpc.c), gntdev_ioctl_dmabuf_exp_from_refs (drivers/xen/gntdev-dmabuf.c), and I think video_usercopy (drivers/media/v4l2-core/v4l2-ioctl.c) also suffer from this same problem. Want to take a stab at those too? :)

-T.J.

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint (spotted by T.J. Mercier and sashiko-bot on v2)
- Add fastrpc fix using the new helper (suggested by T.J. Mercier)

Baineng Shou (2):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds

 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 drivers/misc/fastrpc.c     | 16 +++-----
 include/linux/dma-buf.h    |  1 +
 4 files changed, 67 insertions(+), 50 deletions(-)

2.34.1

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. dma_buf_fd_install() publishes the fd and emits the
trace_dma_buf_fd tracepoint -- from here on the ioctl cannot fail.

A new dma_buf_fd_install() helper is introduced in dma-buf.c to wrap fd_install() together with the DMA_BUF_TRACE() call, preserving the export tracing that dma_buf_fd() provides. dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 include/linux/dma-buf.h    |  1 +
 3 files changed, 61 insertions(+), 40 deletions(-)

index d504c636dc29..4c9add51f9ef 100644
@@ -803,6 +803,26 @@ int dma_buf_fd(struct dma_buf *dmabuf, int flags)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
 
+/**
+ * dma_buf_fd_install - install a reserved fd for a dma-buf
+ * @dmabuf:	[in]	pointer to dma_buf
+ * @fd:		[in]	fd reserved with get_unused_fd_flags()
+ *
+ * Publishes a previously reserved fd into the caller's fd table.
+ * Must only be called after all fallible work (e.g. copy_to_user)
+ * has succeeded, as it cannot be undone safely once called.
+ *
+ * The caller is responsible for having emitted the trace event
+ * (via dma_buf_fd() or get_unused_fd_flags() + this function)
+ * before calling this.
+ */
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd)
+{
+	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
+	fd_install(fd, dmabuf->file);
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_fd_install, "DMA_BUF");
+
 /**
  * dma_buf_get - returns the struct dma_buf related to an fd
  * @fd:	[in]	fd associated with the struct dma_buf to be returned

index d1203da56fc5..d15b2b31d3c9 100644
@@ -567,6 +567,7 @@ void dma_buf_unpin(struct dma_buf_attachment *attach);
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd);
 struct dma_buf *dma_buf_get(int fd);
 void dma_buf_put(struct dma_buf *dmabuf);
 

2.34.1

...

My 2c:

The other option is just to leave it as a 'problem for user space'. No reasonable program is going to handle the EFAULT return by doing anything other than exiting.
Even getting an EFAULT is really an indication that the application is already in a real mess - most likely with a badly corrupted heap.

Anything else leaves error recovery code in the kernel that is pretty much never executed and open to a variety of bugs.
While the recovery here is probably ok, there are some sockopt calls where it is all more complicated.

David

Hi David,

Thanks for the feedback.

The concern is not just about the EFAULT return — it's about the race window between fd_install() and copy_to_user(). Once fd_install() returns, the fd is immediately observable by other threads in the same process (via /proc/self/fd, SCM_RIGHTS, etc.), even before copy_to_user() has a chance to fail. The triggering condition is a deliberate mprotect() flip, not a corrupted heap.

The fix itself is small and follows the standard kernel idiom:
get_unused_fd_flags() reserves the fd without publishing it, so the window between reservation and install is entirely under kernel control.

Baineng

That is what makes doing the close wrong.
But that is a program aggressively trying to hit the timing window, not a normal program that has managed to pass an invalid pointer. The most likely reason for a real program passing an invalid pointer is a corrupted heap (assuming the stupid coding errors are fixed).

It is really no different from the sockopt code that receives SCM_RIGHTS messages.
In that case once you've removed the FILE from the socket (or similar) you really don't want to have to put it back because the write to the sockopt buffer or length field fails.
The chance of correctly reverting the kernel state is small - and won't be tested.

David

Hi David

The rollback here is the simplest possible case (put_unused_fd + dma_buf_put), not the complex sockopt scenario you describe.

Baineng

Hello Baineng,

Thank you for the patch.

While discussing with T J, we realised that adding a selftest for this would also be helpful. Could you please re-submit with a selftest added?

Thanks and best regards,
Sumit.

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip between copy_from_user and copy_to_user) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/
v4: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 +++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 115 +++++++++++++++++-
 6 files changed, 182 insertions(+), 52 deletions(-)

2.34.1

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. dma_buf_fd_install() publishes the fd and emits the
trace_dma_buf_fd tracepoint -- from here on the ioctl cannot fail.

A new dma_buf_fd_install() helper is introduced in dma-buf.c to wrap fd_install() together with the DMA_BUF_TRACE() call, preserving the export tracing that dma_buf_fd() provides. dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 include/linux/dma-buf.h    |  1 +
 3 files changed, 61 insertions(+), 40 deletions(-)

index d504c636dc29..4c9add51f9ef 100644
@@ -803,6 +803,26 @@ int dma_buf_fd(struct dma_buf *dmabuf, int flags)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
 
+/**
+ * dma_buf_fd_install - install a reserved fd for a dma-buf
+ * @dmabuf:	[in]	pointer to dma_buf
+ * @fd:		[in]	fd reserved with get_unused_fd_flags()
+ *
+ * Publishes a previously reserved fd into the caller's fd table.
+ * Must only be called after all fallible work (e.g. copy_to_user)
+ * has succeeded, as it cannot be undone safely once called.
+ *
+ * The caller is responsible for having emitted the trace event
+ * (via dma_buf_fd() or get_unused_fd_flags() + this function)
+ * before calling this.
+ */
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd)
+{
+	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
+	fd_install(fd, dmabuf->file);
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_fd_install, "DMA_BUF");
+
 /**
  * dma_buf_get - returns the struct dma_buf related to an fd
  * @fd:	[in]	fd associated with the struct dma_buf to be returned

index d1203da56fc5..d15b2b31d3c9 100644
@@ -567,6 +567,7 @@ void dma_buf_unpin(struct dma_buf_attachment *attach);
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd);
 struct dma_buf *dma_buf_get(int fd);
 void dma_buf_put(struct dma_buf *dmabuf);
 

2.34.1

fastrpc_ioctl_alloc_dmabuf() calls dma_buf_fd() which installs the fd into the caller's fd table before copy_to_user() copies the fd number back to userspace. If copy_to_user() fails, the fd is already visible to other threads in the same process but the ioctl returns -EFAULT. The existing comment in the code even acknowledges the problem:

"The usercopy failed, but we can't do much about it, as dma_buf_fd()
already called fd_install()..."

Now that dma_buf_fd_install() is available (introduced to fix the same issue in dma-heap), apply the same pattern here: reserve the fd with get_unused_fd_flags(), attempt copy_to_user(), and only on success call dma_buf_fd_install() to publish it atomically with the tracepoint. On copy_to_user() failure, put_unused_fd() and dma_buf_put() cleanly unwind without any user-visible side effects.

---
 drivers/misc/fastrpc.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

index f3a49384586d..c5143cd25767 100644
@@ -1709,24 +1709,20 @@ static int fastrpc_dmabuf_alloc(struct fastrpc_user *fl, char __user *argp)
 		return err;
 	}
 
-	bp.fd = dma_buf_fd(buf->dmabuf, O_ACCMODE);
+	bp.fd = get_unused_fd_flags(O_ACCMODE);
 	if (bp.fd < 0) {
 		dma_buf_put(buf->dmabuf);
-		return -EINVAL;
+		return bp.fd;
 	}
 
 	if (copy_to_user(argp, &bp, sizeof(bp))) {
-		/*
-		 * The usercopy failed, but we can't do much about it, as
-		 * dma_buf_fd() already called fd_install() and made the
-		 * file descriptor accessible for the current process. It
-		 * might already be closed and dmabuf no longer valid when
-		 * we reach this point. Therefore "leak" the fd and rely on
-		 * the process exit path to do any required cleanup.
-		 */
+		put_unused_fd(bp.fd);
+		dma_buf_put(buf->dmabuf);
 		return -EFAULT;
 	}
 
+	dma_buf_fd_install(buf->dmabuf, bp.fd);
+
 	return 0;
 }
 

2.34.1

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip between copy_from_user and copy_to_user) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/
v4: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 +++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 115 +++++++++++++++++-
 6 files changed, 182 insertions(+), 52 deletions(-)

2.34.1

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. dma_buf_fd_install() publishes the fd and emits the
trace_dma_buf_fd tracepoint -- from here on the ioctl cannot fail.

A new dma_buf_fd_install() helper is introduced in dma-buf.c to wrap fd_install() together with the DMA_BUF_TRACE() call, preserving the export tracing that dma_buf_fd() provides. dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 include/linux/dma-buf.h    |  1 +
 3 files changed, 61 insertions(+), 40 deletions(-)

index d504c636dc29..4c9add51f9ef 100644
@@ -803,6 +803,26 @@ int dma_buf_fd(struct dma_buf *dmabuf, int flags)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
 
+/**
+ * dma_buf_fd_install - install a reserved fd for a dma-buf
+ * @dmabuf:	[in]	pointer to dma_buf
+ * @fd:		[in]	fd reserved with get_unused_fd_flags()
+ *
+ * Publishes a previously reserved fd into the caller's fd table.
+ * Must only be called after all fallible work (e.g. copy_to_user)
+ * has succeeded, as it cannot be undone safely once called.
+ *
+ * The caller is responsible for having emitted the trace event
+ * (via dma_buf_fd() or get_unused_fd_flags() + this function)
+ * before calling this.
+ */
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd)
+{
+	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
+	fd_install(fd, dmabuf->file);
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_fd_install, "DMA_BUF");
+
 /**
  * dma_buf_get - returns the struct dma_buf related to an fd
  * @fd:	[in]	fd associated with the struct dma_buf to be returned

index d1203da56fc5..d15b2b31d3c9 100644
@@ -567,6 +567,7 @@ void dma_buf_unpin(struct dma_buf_attachment *attach);
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd);
 struct dma_buf *dma_buf_get(int fd);
 void dma_buf_put(struct dma_buf *dmabuf);
 

2.34.1

fastrpc_ioctl_alloc_dmabuf() calls dma_buf_fd() which installs the fd into the caller's fd table before copy_to_user() copies the fd number back to userspace. If copy_to_user() fails, the fd is already visible to other threads in the same process but the ioctl returns -EFAULT. The existing comment in the code even acknowledges the problem:

"The usercopy failed, but we can't do much about it, as dma_buf_fd()
already called fd_install()..."

Now that dma_buf_fd_install() is available (introduced to fix the same issue in dma-heap), apply the same pattern here: reserve the fd with get_unused_fd_flags(), attempt copy_to_user(), and only on success call dma_buf_fd_install() to publish it atomically with the tracepoint. On copy_to_user() failure, put_unused_fd() and dma_buf_put() cleanly unwind without any user-visible side effects.

---
 drivers/misc/fastrpc.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

index f3a49384586d..c5143cd25767 100644
@@ -1709,24 +1709,20 @@ static int fastrpc_dmabuf_alloc(struct fastrpc_user *fl, char __user *argp)
 		return err;
 	}
 
-	bp.fd = dma_buf_fd(buf->dmabuf, O_ACCMODE);
+	bp.fd = get_unused_fd_flags(O_ACCMODE);
 	if (bp.fd < 0) {
 		dma_buf_put(buf->dmabuf);
-		return -EINVAL;
+		return bp.fd;
 	}
 
 	if (copy_to_user(argp, &bp, sizeof(bp))) {
-		/*
-		 * The usercopy failed, but we can't do much about it, as
-		 * dma_buf_fd() already called fd_install() and made the
-		 * file descriptor accessible for the current process. It
-		 * might already be closed and dmabuf no longer valid when
-		 * we reach this point. Therefore "leak" the fd and rely on
-		 * the process exit path to do any required cleanup.
-		 */
+		put_unused_fd(bp.fd);
+		dma_buf_put(buf->dmabuf);
 		return -EFAULT;
 	}
 
+	dma_buf_fd_install(buf->dmabuf, bp.fd);
+
 	return 0;
 }
 

2.34.1

drm_gem_prime_handle_to_fd() open-codes fd reservation and install using get_unused_fd_flags() + fd_install() directly. This bypasses the DMA_BUF_TRACE() call that dma_buf_fd() emits, so observability tools relying on the trace_dma_buf_fd tracepoint silently miss all DRM PRIME exports.

Replace the bare fd_install() with dma_buf_fd_install(), which wraps fd_install() together with DMA_BUF_TRACE(), restoring full tracepoint coverage. No functional change; the fd lifecycle (get_unused_fd_flags → work → install) is already correct.

Note: this patch depends on dma_buf_fd_install() introduced in "dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds" [1].

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/gpu/drm/drm_prime.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 9b44c78cd77f..fe3436d1235d 100644
@@ -524,7 +524,7 @@ int drm_gem_prime_handle_to_fd(struct drm_device *dev,
 		return PTR_ERR(dmabuf);
 	}
 
-	fd_install(fd, dmabuf->file);
+	dma_buf_fd_install(dmabuf, fd);
 	*prime_fd = fd;
 	return 0;
 }

2.34.1

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip before the ioctl) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v6:
- Rework the selftest (patch 4) per review: extract a count_open_fds() helper, fix the copy_from_user() comment, fail (not skip) when the ioctl does not return -1, drop the bogus mprotect-race mention, and reword the result message.

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 ++++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++-------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 6 files changed, 180 insertions(+), 52 deletions(-)

2.34.1

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip before the ioctl) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v6:
- Rework the selftest (patch 4) per review: extract a count_open_fds() helper, fix the copy_from_user() comment, fail (not skip) when the ioctl does not return -1, drop the bogus mprotect-race mention, and reword the result message.

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 ++++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++-------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 6 files changed, 180 insertions(+), 52 deletions(-)

2.34.1

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. dma_buf_fd_install() publishes the fd and emits the
trace_dma_buf_fd tracepoint -- from here on the ioctl cannot fail.

A new dma_buf_fd_install() helper is introduced in dma-buf.c to wrap fd_install() together with the DMA_BUF_TRACE() call, preserving the export tracing that dma_buf_fd() provides. dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 include/linux/dma-buf.h    |  1 +
 3 files changed, 61 insertions(+), 40 deletions(-)

index d504c636dc29..4c9add51f9ef 100644
@@ -803,6 +803,26 @@ int dma_buf_fd(struct dma_buf *dmabuf, int flags)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
 
+/**
+ * dma_buf_fd_install - install a reserved fd for a dma-buf
+ * @dmabuf:	[in]	pointer to dma_buf
+ * @fd:		[in]	fd reserved with get_unused_fd_flags()
+ *
+ * Publishes a previously reserved fd into the caller's fd table.
+ * Must only be called after all fallible work (e.g. copy_to_user)
+ * has succeeded, as it cannot be undone safely once called.
+ *
+ * The caller is responsible for having emitted the trace event
+ * (via dma_buf_fd() or get_unused_fd_flags() + this function)
+ * before calling this.
+ */
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd)
+{
+	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
+	fd_install(fd, dmabuf->file);
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_fd_install, "DMA_BUF");
+
 /**
  * dma_buf_get - returns the struct dma_buf related to an fd
  * @fd:	[in]	fd associated with the struct dma_buf to be returned

index d1203da56fc5..d15b2b31d3c9 100644
@@ -567,6 +567,7 @@ void dma_buf_unpin(struct dma_buf_attachment *attach);
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd);
 struct dma_buf *dma_buf_get(int fd);
 void dma_buf_put(struct dma_buf *dmabuf);
 

2.34.1

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip before the ioctl) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v6:
- Rework the selftest (patch 4) per review: extract a count_open_fds() helper, fix the copy_from_user() comment, fail (not skip) when the ioctl does not return -1, drop the bogus mprotect-race mention, and reword the result message.

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 ++++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++-------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 6 files changed, 180 insertions(+), 52 deletions(-)

2.34.1

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip before the ioctl) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v6:
- Rework the selftest (patch 4) per review: extract a count_open_fds() helper, fix the copy_from_user() comment, fail (not skip) when the ioctl does not return -1, drop the bogus mprotect-race mention, and reword the result message.

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 ++++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++-------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 6 files changed, 180 insertions(+), 52 deletions(-)

2.34.1

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip before the ioctl) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v6:
- Rework the selftest (patch 4) per review: extract a count_open_fds() helper, fix the copy_from_user() comment, fail (not skip) when the ioctl does not return -1, drop the bogus mprotect-race mention, and reword the result message.

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 ++++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++-------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 6 files changed, 180 insertions(+), 52 deletions(-)

2.34.1

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. dma_buf_fd_install() publishes the fd and emits the
trace_dma_buf_fd tracepoint -- from here on the ioctl cannot fail.

A new dma_buf_fd_install() helper is introduced in dma-buf.c to wrap fd_install() together with the DMA_BUF_TRACE() call, preserving the export tracing that dma_buf_fd() provides. dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 include/linux/dma-buf.h    |  1 +
 3 files changed, 61 insertions(+), 40 deletions(-)

index d504c636dc29..4c9add51f9ef 100644
@@ -803,6 +803,26 @@ int dma_buf_fd(struct dma_buf *dmabuf, int flags)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
 
+/**
+ * dma_buf_fd_install - install a reserved fd for a dma-buf
+ * @dmabuf:	[in]	pointer to dma_buf
+ * @fd:		[in]	fd reserved with get_unused_fd_flags()
+ *
+ * Publishes a previously reserved fd into the caller's fd table.
+ * Must only be called after all fallible work (e.g. copy_to_user)
+ * has succeeded, as it cannot be undone safely once called.
+ *
+ * The caller is responsible for having emitted the trace event
+ * (via dma_buf_fd() or get_unused_fd_flags() + this function)
+ * before calling this.
+ */
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd)
+{
+	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
+	fd_install(fd, dmabuf->file);
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_fd_install, "DMA_BUF");
+
 /**
  * dma_buf_get - returns the struct dma_buf related to an fd
  * @fd:	[in]	fd associated with the struct dma_buf to be returned

index d1203da56fc5..d15b2b31d3c9 100644
@@ -567,6 +567,7 @@ void dma_buf_unpin(struct dma_buf_attachment *attach);
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd);
 struct dma_buf *dma_buf_get(int fd);
 void dma_buf_put(struct dma_buf *dmabuf);
 

2.34.1

fastrpc_ioctl_alloc_dmabuf() calls dma_buf_fd() which installs the fd into the caller's fd table before copy_to_user() copies the fd number back to userspace. If copy_to_user() fails, the fd is already visible to other threads in the same process but the ioctl returns -EFAULT. The existing comment in the code even acknowledges the problem:

"The usercopy failed, but we can't do much about it, as dma_buf_fd()
already called fd_install()..."

Now that dma_buf_fd_install() is available (introduced to fix the same issue in dma-heap), apply the same pattern here: reserve the fd with get_unused_fd_flags(), attempt copy_to_user(), and only on success call dma_buf_fd_install() to publish it atomically with the tracepoint. On copy_to_user() failure, put_unused_fd() and dma_buf_put() cleanly unwind without any user-visible side effects.

---
 drivers/misc/fastrpc.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

index f3a49384586d..c5143cd25767 100644
@@ -1709,24 +1709,20 @@ static int fastrpc_dmabuf_alloc(struct fastrpc_user *fl, char __user *argp)
 		return err;
 	}
 
-	bp.fd = dma_buf_fd(buf->dmabuf, O_ACCMODE);
+	bp.fd = get_unused_fd_flags(O_ACCMODE);
 	if (bp.fd < 0) {
 		dma_buf_put(buf->dmabuf);
-		return -EINVAL;
+		return bp.fd;
 	}
 
 	if (copy_to_user(argp, &bp, sizeof(bp))) {
-		/*
-		 * The usercopy failed, but we can't do much about it, as
-		 * dma_buf_fd() already called fd_install() and made the
-		 * file descriptor accessible for the current process. It
-		 * might already be closed and dmabuf no longer valid when
-		 * we reach this point. Therefore "leak" the fd and rely on
-		 * the process exit path to do any required cleanup.
-		 */
+		put_unused_fd(bp.fd);
+		dma_buf_put(buf->dmabuf);
 		return -EFAULT;
 	}
 
+	dma_buf_fd_install(buf->dmabuf, bp.fd);
+
 	return 0;
 }
 

2.34.1

drm_gem_prime_handle_to_fd() open-codes fd reservation and install using get_unused_fd_flags() + fd_install() directly. This bypasses the DMA_BUF_TRACE() call that dma_buf_fd() emits, so observability tools relying on the trace_dma_buf_fd tracepoint silently miss all DRM PRIME exports.

Replace the bare fd_install() with dma_buf_fd_install(), which wraps fd_install() together with DMA_BUF_TRACE(), restoring full tracepoint coverage. No functional change; the fd lifecycle (get_unused_fd_flags → work → install) is already correct.

Note: this patch depends on dma_buf_fd_install() introduced in "dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds" [1].

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/gpu/drm/drm_prime.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 9b44c78cd77f..fe3436d1235d 100644
@@ -524,7 +524,7 @@ int drm_gem_prime_handle_to_fd(struct drm_device *dev,
 		return PTR_ERR(dmabuf);
 	}
 
-	fd_install(fd, dmabuf->file);
+	dma_buf_fd_install(dmabuf, fd);
 	*prime_fd = fd;
 	return 0;
 }

2.34.1

Add a test case that verifies no file descriptor is leaked when DMA_HEAP_IOCTL_ALLOC succeeds internally but copy_to_user() fails to deliver the fd number back to userspace.

The failure is triggered by placing the ioctl argument in a private anonymous page and flipping it to PROT_READ (via mprotect) between the kernel's copy_from_user() and copy_to_user() calls. With the buggy kernel the ioctl returns -EFAULT but leaves an extra open fd in the process's fd table; with the fixed kernel the fd count is unchanged.

This serves as a regression test for:
"dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds"

---
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 1 file changed, 112 insertions(+), 1 deletion(-)

index fc9694fc4e89..1d49df671919 100644
@@ -390,6 +390,116 @@ static void test_alloc_errors(char *heap_name)
 	close(heap_fd);
 }
 
+/*
+ * count_open_fds - return the number of open file descriptors.
+ *
+ * The fd opened by opendir() itself is counted, but since it is opened
+ * and closed within each call, it cancels out when comparing two counts.
+ * Returns -1 on error.
+ */
+static int count_open_fds(void)
+{
+	DIR *d = opendir("/proc/self/fd");
+	struct dirent *de;
+	int count = 0;
+
+	if (!d)
+		return -1;
+
+	while ((de = readdir(d)))
+		if (de->d_name[0] != '.')
+			count++;
+	closedir(d);
+	return count;
+}
+
+/*
+ * test_alloc_no_fd_leak_on_efault - verify no fd is leaked when
+ * copy_to_user() fails during DMA_HEAP_IOCTL_ALLOC.
+ *
+ * The bug: dma_buf_fd() called fd_install() before copy_to_user().
+ * If copy_to_user() then failed (e.g. via mprotect), the fd was
+ * silently installed in the fd table but never returned to userspace.
+ *
+ * The fix: reserve the fd with get_unused_fd_flags() first, attempt
+ * copy_to_user(), and only call fd_install() on success.
+ *
+ * We trigger the failure by placing the ioctl argument in a private
+ * anonymous page and flipping it to PROT_READ before the ioctl.
+ * Inside the kernel, copy_from_user() reads from the page (reads are
+ * allowed under PROT_READ, so it succeeds), but copy_to_user() that
+ * writes the fd number back faults, returning -EFAULT.  We then
+ * count open file descriptors before and after; with the bug an extra
+ * fd is left in the table.
+ */
+static void test_alloc_no_fd_leak_on_efault(char *heap_name)
+{
+	int heap_fd = -1;
+	int fd_before, fd_after;
+	int ret;
+	long page_size;
+	struct dma_heap_allocation_data *req;
+
+	ksft_print_msg("Testing fd leak when copy_to_user() fails:\n");
+
+	heap_fd = dmabuf_heap_open(heap_name);
+
+	page_size = sysconf(_SC_PAGESIZE);
+
+	/*
+	 * Place the ioctl argument in its own private anonymous page so
+	 * we can flip its protection independently.
+	 */
+	req = mmap(NULL, page_size, PROT_READ | PROT_WRITE,
+		   MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	if (req == MAP_FAILED) {
+		ksft_test_result_fail("mmap failed: %s\n", strerror(errno));
+		goto out;
+	}
+
+	memset(req, 0, sizeof(*req));
+	req->len      = page_size;
+	req->fd_flags = O_RDWR | O_CLOEXEC;
+
+	fd_before = count_open_fds();
+	if (fd_before < 0) {
+		ksft_test_result_fail("count_open_fds: %s\n", strerror(errno));
+		munmap(req, page_size);
+		goto out;
+	}
+
+	/*
+	 * Make the page read-only so copy_to_user() will fault.  The
+	 * ioctl must fail with -1; if it returns success the test setup
+	 * is broken (mprotect is synchronous, so there is no race).
+	 */
+	mprotect(req, page_size, PROT_READ);
+
+	ret = ioctl(heap_fd, DMA_HEAP_IOCTL_ALLOC, req);
+
+	/* Re-allow writes so munmap can clean up */
+	mprotect(req, page_size, PROT_READ | PROT_WRITE);
+	munmap(req, page_size);
+
+	if (ret != -1) {
+		ksft_test_result_fail("ioctl returned %d, expected -1 EFAULT\n",
+				      ret);
+		goto out;
+	}
+
+	fd_after = count_open_fds();
+	if (fd_after < 0) {
+		ksft_test_result_fail("count_open_fds: %s\n", strerror(errno));
+		goto out;
+	}
+
+	ksft_test_result(fd_before == fd_after,
+			 "fd leak on EFAULT: before=%d after=%d\n",
+			 fd_before, fd_after);
+out:
+	close(heap_fd);
+}
+
 static int numer_of_heaps(void)
 {
 	DIR *d = opendir(DEVPATH);

@@ -420,7 +530,7 @@ int main(void)
 		return KSFT_SKIP;
 	}
 
-	ksft_set_plan(11 * numer_of_heaps());
+	ksft_set_plan(12 * numer_of_heaps());
 
 	while ((dir = readdir(d))) {
 		if (!strncmp(dir->d_name, ".", 2))

@@ -435,6 +545,7 @@ int main(void)
 		test_alloc_zeroed(dir->d_name, ONE_MEG);
 		test_alloc_compat(dir->d_name);
 		test_alloc_errors(dir->d_name);
+		test_alloc_no_fd_leak_on_efault(dir->d_name);
 	}
 	closedir(d);
 

2.34.1

Hello Baineng,

Thank you for the series - please feel free to add my

Best,
Sumit.

Several drivers call dma_buf_fd() — which internally calls fd_install() — before copy_to_user() returns the fd number to userspace. If copy_to_user() fails, the fd is already published in the caller's fd table but the ioctl returns an error, so userspace never learns the fd number. Worse, the window between fd_install() and copy_to_user() allows other threads to observe and manipulate the fd (dup, close, SCM_RIGHTS), making any "close it on the failure path" fix unsafe.

The fix is to split the allocation into three steps: reserve an fd with get_unused_fd_flags() (not yet visible to other threads), do copy_to_user(), and only then publish the fd with fd_install() via the new dma_buf_fd_install() helper. On copy_to_user() failure, put_unused_fd() + dma_buf_put() cleanly unwind with no user-visible side effects.

Patch 1 introduces dma_buf_fd_install() in dma-buf.c (wrapping fd_install() together with the DMA_BUF_TRACE call to preserve export tracing) and applies the fix to dma-heap.

Patch 2 applies the same fix to fastrpc, which even had a comment acknowledging the problem could not be fixed before.

Patch 3 replaces the bare fd_install() in drm_gem_prime_handle_to_fd() with dma_buf_fd_install() to restore tracepoint coverage for DRM PRIME exports (suggested by Christian König).

Patch 4 adds a selftest to tools/testing/selftests/dmabuf-heaps/ that reproduces the fd-leak scenario (mprotect flip before the ioctl) and verifies the fd count is unchanged after a failed ioctl (suggested by Sumit Semwal).

v1: https://lore.kernel.org/dri-devel/[email protected]/
v2: https://lore.kernel.org/dri-devel/[email protected]/
v3: https://lore.kernel.org/dri-devel/[email protected]/
v5: https://lore.kernel.org/dri-devel/[email protected]/
v6: https://lore.kernel.org/dri-devel/[email protected]/

Changes in v7:
- Add Reviewed-by: T.J. Mercier to patch 4 (selftest).
- Add Acked-by: Sumit Semwal to the whole series.

Changes in v6:
- Rework the selftest (patch 4) per review: extract a count_open_fds() helper, fix the copy_from_user() comment, fail (not skip) when the ioctl does not return -1, drop the bogus mprotect-race mention, and reword the result message.

Changes in v5:
- Add selftest (patch 4) reproducing the fd-leak scenario (Sumit Semwal)

Changes in v4:
- Add patch 3: drm/prime: use dma_buf_fd_install() (Christian König)
- Add Acked-by: Christian König to patches 1 and 2

Changes in v3:
- Split into two patches (dma-heap + fastrpc separately)
- Add dma_buf_fd_install() to preserve trace_dma_buf_fd tracepoint
- Add fastrpc fix using the new helper (T.J. Mercier)

Baineng Shou (4):
dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds
misc: fastrpc: don't publish fd before copy_to_user() succeeds drm/prime: use dma_buf_fd_install() to preserve export tracing
selftests: dmabuf-heaps: add fd-leak-on-EFAULT regression test

 drivers/dma-buf/dma-buf.c                     |  20 ++++
 drivers/dma-buf/dma-heap.c                    |  80 ++++++-------
 drivers/gpu/drm/drm_prime.c                   |   2 +-
 drivers/misc/fastrpc.c                        |  16 +--
 include/linux/dma-buf.h                       |   1 +
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 6 files changed, 180 insertions(+), 52 deletions(-)

2.34.1

DMA_HEAP_IOCTL_ALLOC allocates a dma-buf and installs an fd into the caller's fd table via dma_buf_fd() -> fd_install() before dma_heap_ioctl() copies the result back to userspace. If the trailing copy_to_user() fails, userspace never learns the fd number, but the fd (and the underlying dma-buf reference) are already visible to other threads in the same process and are leaked for the lifetime of the process.

The obvious "close it on the failure path" fix is unsafe: once fd_install() has run, another thread can already dup() the fd, send it via SCM_RIGHTS, or close() it and let its number be reused, so a subsequent close_fd() from the ioctl path can operate on an unrelated file. This was pointed out by Christian König on v1 [1].

Restructure the allocation path so that fd_install() is the last, unfailable step of a successful ioctl:

1. heap->ops->allocate() creates the dma_buf.
2. get_unused_fd_flags() reserves an fd number in the caller's
fd table without publishing it, so no other thread can observe it.
3. copy_to_user() delivers the fd number to userspace;
on failure the fd is returned with put_unused_fd() and the dma_buf reference is dropped with dma_buf_put(), leaving no user- visible state behind.
4. dma_buf_fd_install() publishes the fd and emits the
trace_dma_buf_fd tracepoint -- from here on the ioctl cannot fail.

A new dma_buf_fd_install() helper is introduced in dma-buf.c to wrap fd_install() together with the DMA_BUF_TRACE() call, preserving the export tracing that dma_buf_fd() provides. dma_heap_ioctl_allocate() is refactored to return the struct dma_buf * directly (returning ERR_PTR on failure) so the caller holds the dmabuf reference across steps 3 and 4.

The failure at step 3 is easily reachable from userspace: pass a struct dma_heap_allocation_data that lives in a page whose protection is flipped to PROT_READ between copy_from_user() and copy_to_user() (e.g. via mprotect()). Before this change each such ioctl leaks one dmabuf fd; after it, the fd table is unchanged on failure and only /dev/dma_heap/<name> remains open.

No UAPI or heap-driver interface change.

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/dma-buf/dma-buf.c  | 20 ++++++++++
 drivers/dma-buf/dma-heap.c | 80 +++++++++++++++++++-------------------
 include/linux/dma-buf.h    |  1 +
 3 files changed, 61 insertions(+), 40 deletions(-)

index d504c636dc29..4c9add51f9ef 100644
@@ -803,6 +803,26 @@ int dma_buf_fd(struct dma_buf *dmabuf, int flags)
 }
 EXPORT_SYMBOL_NS_GPL(dma_buf_fd, "DMA_BUF");
 
+/**
+ * dma_buf_fd_install - install a reserved fd for a dma-buf
+ * @dmabuf:	[in]	pointer to dma_buf
+ * @fd:		[in]	fd reserved with get_unused_fd_flags()
+ *
+ * Publishes a previously reserved fd into the caller's fd table.
+ * Must only be called after all fallible work (e.g. copy_to_user)
+ * has succeeded, as it cannot be undone safely once called.
+ *
+ * The caller is responsible for having emitted the trace event
+ * (via dma_buf_fd() or get_unused_fd_flags() + this function)
+ * before calling this.
+ */
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd)
+{
+	DMA_BUF_TRACE(trace_dma_buf_fd, dmabuf, fd);
+	fd_install(fd, dmabuf->file);
+}
+EXPORT_SYMBOL_NS_GPL(dma_buf_fd_install, "DMA_BUF");
+
 /**
  * dma_buf_get - returns the struct dma_buf related to an fd
  * @fd:	[in]	fd associated with the struct dma_buf to be returned

index d1203da56fc5..d15b2b31d3c9 100644
@@ -567,6 +567,7 @@ void dma_buf_unpin(struct dma_buf_attachment *attach);
 struct dma_buf *dma_buf_export(const struct dma_buf_export_info *exp_info);
 
 int dma_buf_fd(struct dma_buf *dmabuf, int flags);
+void dma_buf_fd_install(struct dma_buf *dmabuf, int fd);
 struct dma_buf *dma_buf_get(int fd);
 void dma_buf_put(struct dma_buf *dmabuf);
 

2.34.1

fastrpc_ioctl_alloc_dmabuf() calls dma_buf_fd() which installs the fd into the caller's fd table before copy_to_user() copies the fd number back to userspace. If copy_to_user() fails, the fd is already visible to other threads in the same process but the ioctl returns -EFAULT. The existing comment in the code even acknowledges the problem:

"The usercopy failed, but we can't do much about it, as dma_buf_fd()
already called fd_install()..."

Now that dma_buf_fd_install() is available (introduced to fix the same issue in dma-heap), apply the same pattern here: reserve the fd with get_unused_fd_flags(), attempt copy_to_user(), and only on success call dma_buf_fd_install() to publish it atomically with the tracepoint. On copy_to_user() failure, put_unused_fd() and dma_buf_put() cleanly unwind without any user-visible side effects.

---
 drivers/misc/fastrpc.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

index f3a49384586d..c5143cd25767 100644
@@ -1709,24 +1709,20 @@ static int fastrpc_dmabuf_alloc(struct fastrpc_user *fl, char __user *argp)
 		return err;
 	}
 
-	bp.fd = dma_buf_fd(buf->dmabuf, O_ACCMODE);
+	bp.fd = get_unused_fd_flags(O_ACCMODE);
 	if (bp.fd < 0) {
 		dma_buf_put(buf->dmabuf);
-		return -EINVAL;
+		return bp.fd;
 	}
 
 	if (copy_to_user(argp, &bp, sizeof(bp))) {
-		/*
-		 * The usercopy failed, but we can't do much about it, as
-		 * dma_buf_fd() already called fd_install() and made the
-		 * file descriptor accessible for the current process. It
-		 * might already be closed and dmabuf no longer valid when
-		 * we reach this point. Therefore "leak" the fd and rely on
-		 * the process exit path to do any required cleanup.
-		 */
+		put_unused_fd(bp.fd);
+		dma_buf_put(buf->dmabuf);
 		return -EFAULT;
 	}
 
+	dma_buf_fd_install(buf->dmabuf, bp.fd);
+
 	return 0;
 }
 

2.34.1

drm_gem_prime_handle_to_fd() open-codes fd reservation and install using get_unused_fd_flags() + fd_install() directly. This bypasses the DMA_BUF_TRACE() call that dma_buf_fd() emits, so observability tools relying on the trace_dma_buf_fd tracepoint silently miss all DRM PRIME exports.

Replace the bare fd_install() with dma_buf_fd_install(), which wraps fd_install() together with DMA_BUF_TRACE(), restoring full tracepoint coverage. No functional change; the fd lifecycle (get_unused_fd_flags → work → install) is already correct.

Note: this patch depends on dma_buf_fd_install() introduced in "dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds" [1].

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/gpu/drm/drm_prime.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 9b44c78cd77f..fe3436d1235d 100644
@@ -524,7 +524,7 @@ int drm_gem_prime_handle_to_fd(struct drm_device *dev,
 		return PTR_ERR(dmabuf);
 	}
 
-	fd_install(fd, dmabuf->file);
+	dma_buf_fd_install(dmabuf, fd);
 	*prime_fd = fd;
 	return 0;
 }

2.34.1

Add a test case that verifies no file descriptor is leaked when DMA_HEAP_IOCTL_ALLOC succeeds internally but copy_to_user() fails to deliver the fd number back to userspace.

The failure is triggered by placing the ioctl argument in a private anonymous page and flipping it to PROT_READ (via mprotect) between the kernel's copy_from_user() and copy_to_user() calls. With the buggy kernel the ioctl returns -EFAULT but leaves an extra open fd in the process's fd table; with the fixed kernel the fd count is unchanged.

This serves as a regression test for:
"dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds"

---
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 113 +++++++++++++++++-
 1 file changed, 112 insertions(+), 1 deletion(-)

index fc9694fc4e89..1d49df671919 100644
@@ -390,6 +390,116 @@ static void test_alloc_errors(char *heap_name)
 	close(heap_fd);
 }
 
+/*
+ * count_open_fds - return the number of open file descriptors.
+ *
+ * The fd opened by opendir() itself is counted, but since it is opened
+ * and closed within each call, it cancels out when comparing two counts.
+ * Returns -1 on error.
+ */
+static int count_open_fds(void)
+{
+	DIR *d = opendir("/proc/self/fd");
+	struct dirent *de;
+	int count = 0;
+
+	if (!d)
+		return -1;
+
+	while ((de = readdir(d)))
+		if (de->d_name[0] != '.')
+			count++;
+	closedir(d);
+	return count;
+}
+
+/*
+ * test_alloc_no_fd_leak_on_efault - verify no fd is leaked when
+ * copy_to_user() fails during DMA_HEAP_IOCTL_ALLOC.
+ *
+ * The bug: dma_buf_fd() called fd_install() before copy_to_user().
+ * If copy_to_user() then failed (e.g. via mprotect), the fd was
+ * silently installed in the fd table but never returned to userspace.
+ *
+ * The fix: reserve the fd with get_unused_fd_flags() first, attempt
+ * copy_to_user(), and only call fd_install() on success.
+ *
+ * We trigger the failure by placing the ioctl argument in a private
+ * anonymous page and flipping it to PROT_READ before the ioctl.
+ * Inside the kernel, copy_from_user() reads from the page (reads are
+ * allowed under PROT_READ, so it succeeds), but copy_to_user() that
+ * writes the fd number back faults, returning -EFAULT.  We then
+ * count open file descriptors before and after; with the bug an extra
+ * fd is left in the table.
+ */
+static void test_alloc_no_fd_leak_on_efault(char *heap_name)
+{
+	int heap_fd = -1;
+	int fd_before, fd_after;
+	int ret;
+	long page_size;
+	struct dma_heap_allocation_data *req;
+
+	ksft_print_msg("Testing fd leak when copy_to_user() fails:\n");
+
+	heap_fd = dmabuf_heap_open(heap_name);
+
+	page_size = sysconf(_SC_PAGESIZE);
+
+	/*
+	 * Place the ioctl argument in its own private anonymous page so
+	 * we can flip its protection independently.
+	 */
+	req = mmap(NULL, page_size, PROT_READ | PROT_WRITE,
+		   MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	if (req == MAP_FAILED) {
+		ksft_test_result_fail("mmap failed: %s\n", strerror(errno));
+		goto out;
+	}
+
+	memset(req, 0, sizeof(*req));
+	req->len      = page_size;
+	req->fd_flags = O_RDWR | O_CLOEXEC;
+
+	fd_before = count_open_fds();
+	if (fd_before < 0) {
+		ksft_test_result_fail("count_open_fds: %s\n", strerror(errno));
+		munmap(req, page_size);
+		goto out;
+	}
+
+	/*
+	 * Make the page read-only so copy_to_user() will fault.  The
+	 * ioctl must fail with -1; if it returns success the test setup
+	 * is broken (mprotect is synchronous, so there is no race).
+	 */
+	mprotect(req, page_size, PROT_READ);
+
+	ret = ioctl(heap_fd, DMA_HEAP_IOCTL_ALLOC, req);
+
+	/* Re-allow writes so munmap can clean up */
+	mprotect(req, page_size, PROT_READ | PROT_WRITE);
+	munmap(req, page_size);
+
+	if (ret != -1) {
+		ksft_test_result_fail("ioctl returned %d, expected -1 EFAULT\n",
+				      ret);
+		goto out;
+	}
+
+	fd_after = count_open_fds();
+	if (fd_after < 0) {
+		ksft_test_result_fail("count_open_fds: %s\n", strerror(errno));
+		goto out;
+	}
+
+	ksft_test_result(fd_before == fd_after,
+			 "fd leak on EFAULT: before=%d after=%d\n",
+			 fd_before, fd_after);
+out:
+	close(heap_fd);
+}
+
 static int numer_of_heaps(void)
 {
 	DIR *d = opendir(DEVPATH);

@@ -420,7 +530,7 @@ int main(void)
 		return KSFT_SKIP;
 	}
 
-	ksft_set_plan(11 * numer_of_heaps());
+	ksft_set_plan(12 * numer_of_heaps());
 
 	while ((dir = readdir(d))) {
 		if (!strncmp(dir->d_name, ".", 2))

@@ -435,6 +545,7 @@ int main(void)
 		test_alloc_zeroed(dir->d_name, ONE_MEG);
 		test_alloc_compat(dir->d_name);
 		test_alloc_errors(dir->d_name);
+		test_alloc_no_fd_leak_on_efault(dir->d_name);
 	}
 	closedir(d);
 

2.34.1

I've gone ahead, added one more rb and pushed the result to drm-misc-fixes.

Thanks for the help,
Christian.

Add a test case that verifies no file descriptor is leaked when DMA_HEAP_IOCTL_ALLOC succeeds internally but copy_to_user() fails to deliver the fd number back to userspace.

The failure is triggered by placing the ioctl argument in a private anonymous page and flipping it to PROT_READ (via mprotect) between the kernel's copy_from_user() and copy_to_user() calls. With the buggy kernel the ioctl returns -EFAULT but leaves an extra open fd in the process's fd table; with the fixed kernel the fd count is unchanged.

This serves as a regression test for:
"dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds"

---
 .../selftests/dmabuf-heaps/dmabuf-heap.c      | 115 +++++++++++++++++-
 1 file changed, 114 insertions(+), 1 deletion(-)

index fc9694fc4e89..bd58e5b06c8b 100644
@@ -390,6 +390,118 @@ static void test_alloc_errors(char *heap_name)
 	close(heap_fd);
 }
 
+/*
+ * test_alloc_no_fd_leak_on_efault - verify no fd is leaked when
+ * copy_to_user() fails during DMA_HEAP_IOCTL_ALLOC.
+ *
+ * The bug: dma_buf_fd() called fd_install() before copy_to_user().
+ * If copy_to_user() then failed (e.g. via mprotect), the fd was
+ * silently installed in the fd table but never returned to userspace.
+ *
+ * The fix: reserve the fd with get_unused_fd_flags() first, attempt
+ * copy_to_user(), and only call fd_install() on success.
+ *
+ * We trigger the failure by placing the ioctl argument in a page,
+ * flipping it to PROT_READ between copy_from_user and copy_to_user,
+ * and counting open file descriptors before and after.
+ */
+static void test_alloc_no_fd_leak_on_efault(char *heap_name)
+{
+	int heap_fd = -1;
+	int fd_before, fd_after;
+	int ret;
+	long page_size;
+	struct dma_heap_allocation_data *req;
+
+	ksft_print_msg("Testing no fd leak when copy_to_user() fails:\n");
+
+	heap_fd = dmabuf_heap_open(heap_name);
+
+	page_size = sysconf(_SC_PAGESIZE);
+
+	/*
+	 * Place the ioctl argument in its own private anonymous page so
+	 * we can flip its protection independently.
+	 */
+	req = mmap(NULL, page_size, PROT_READ | PROT_WRITE,
+		   MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+	if (req == MAP_FAILED) {
+		ksft_test_result_fail("mmap failed: %s\n", strerror(errno));
+		goto out;
+	}
+
+	memset(req, 0, sizeof(*req));
+	req->len      = page_size;
+	req->fd_flags = O_RDWR | O_CLOEXEC;
+
+	/* Count open fds before the ioctl */
+	fd_before = 0;
+	{
+		DIR *d = opendir("/proc/self/fd");
+		struct dirent *de;
+
+		if (!d) {
+			ksft_test_result_fail("opendir /proc/self/fd: %s\n",
+					      strerror(errno));
+			munmap(req, page_size);
+			goto out;
+		}
+		while ((de = readdir(d)))
+			if (de->d_name[0] != '.')
+				fd_before++;
+		closedir(d);
+		/* subtract the fd opened by opendir itself */
+	}
+
+	/*
+	 * Make the page read-only: copy_from_user() in the kernel will
+	 * still succeed (it already ran), but copy_to_user() that writes
+	 * the fd number back will fault.
+	 */
+	mprotect(req, page_size, PROT_READ);
+
+	ret = ioctl(heap_fd, DMA_HEAP_IOCTL_ALLOC, req);
+
+	/* Re-allow writes so munmap can clean up */
+	mprotect(req, page_size, PROT_READ | PROT_WRITE);
+	munmap(req, page_size);
+
+	if (ret != -1 || errno != EFAULT) {
+		/*
+		 * If the ioctl didn't fail with EFAULT, either the kernel
+		 * handled it differently or mprotect raced.  Skip rather
+		 * than giving a false pass/fail.
+		 */
+		ksft_test_result_skip(
+			"ioctl did not return EFAULT (ret=%d errno=%d), skipping\n",
+			ret, errno);
+		goto out;
+	}
+
+	/* Count open fds after the failed ioctl */
+	fd_after = 0;
+	{
+		DIR *d = opendir("/proc/self/fd");
+		struct dirent *de;
+
+		if (!d) {
+			ksft_test_result_fail("opendir /proc/self/fd: %s\n",
+					      strerror(errno));
+			goto out;
+		}
+		while ((de = readdir(d)))
+			if (de->d_name[0] != '.')
+				fd_after++;
+		closedir(d);
+	}
+
+	ksft_test_result(fd_before == fd_after,
+			 "no fd leak on EFAULT: before=%d after=%d\n",
+			 fd_before, fd_after);
+out:
+	close(heap_fd);
+}
+
 static int numer_of_heaps(void)
 {
 	DIR *d = opendir(DEVPATH);

@@ -420,7 +532,7 @@ int main(void)
 		return KSFT_SKIP;
 	}
 
-	ksft_set_plan(11 * numer_of_heaps());
+	ksft_set_plan(12 * numer_of_heaps());
 
 	while ((dir = readdir(d))) {
 		if (!strncmp(dir->d_name, ".", 2))

@@ -435,6 +547,7 @@ int main(void)
 		test_alloc_zeroed(dir->d_name, ONE_MEG);
 		test_alloc_compat(dir->d_name);
 		test_alloc_errors(dir->d_name);
+		test_alloc_no_fd_leak_on_efault(dir->d_name);
 	}
 	closedir(d);
 

2.34.1

fastrpc_ioctl_alloc_dmabuf() calls dma_buf_fd() which installs the fd into the caller's fd table before copy_to_user() copies the fd number back to userspace. If copy_to_user() fails, the fd is already visible to other threads in the same process but the ioctl returns -EFAULT. The existing comment in the code even acknowledges the problem:

"The usercopy failed, but we can't do much about it, as dma_buf_fd()
already called fd_install()..."

Now that dma_buf_fd_install() is available (introduced to fix the same issue in dma-heap), apply the same pattern here: reserve the fd with get_unused_fd_flags(), attempt copy_to_user(), and only on success call dma_buf_fd_install() to publish it atomically with the tracepoint. On copy_to_user() failure, put_unused_fd() and dma_buf_put() cleanly unwind without any user-visible side effects.

---
 drivers/misc/fastrpc.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

index f3a49384586d..c5143cd25767 100644
@@ -1709,24 +1709,20 @@ static int fastrpc_dmabuf_alloc(struct fastrpc_user *fl, char __user *argp)
 		return err;
 	}
 
-	bp.fd = dma_buf_fd(buf->dmabuf, O_ACCMODE);
+	bp.fd = get_unused_fd_flags(O_ACCMODE);
 	if (bp.fd < 0) {
 		dma_buf_put(buf->dmabuf);
-		return -EINVAL;
+		return bp.fd;
 	}
 
 	if (copy_to_user(argp, &bp, sizeof(bp))) {
-		/*
-		 * The usercopy failed, but we can't do much about it, as
-		 * dma_buf_fd() already called fd_install() and made the
-		 * file descriptor accessible for the current process. It
-		 * might already be closed and dmabuf no longer valid when
-		 * we reach this point. Therefore "leak" the fd and rely on
-		 * the process exit path to do any required cleanup.
-		 */
+		put_unused_fd(bp.fd);
+		dma_buf_put(buf->dmabuf);
 		return -EFAULT;
 	}
 
+	dma_buf_fd_install(buf->dmabuf, bp.fd);
+
 	return 0;
 }
 

2.34.1

drivers/gpu/drm/drm_prime.c is also using fd_install() of a DMA-buf file descriptor manually.

Would be nice if we could us the new dma_buf_fd_install() for tracing here as well.

Apart from that feel free to add Acked-by: Christian König <[email protected]> to the whole series.

Regards,
Christian.

drm_gem_prime_handle_to_fd() open-codes fd reservation and install using get_unused_fd_flags() + fd_install() directly. This bypasses the DMA_BUF_TRACE() call that dma_buf_fd() emits, so observability tools relying on the trace_dma_buf_fd tracepoint silently miss all DRM PRIME exports.

Replace the bare fd_install() with dma_buf_fd_install(), which wraps fd_install() together with DMA_BUF_TRACE(), restoring full tracepoint coverage. No functional change; the fd lifecycle (get_unused_fd_flags → work → install) is already correct.

Note: this patch depends on dma_buf_fd_install() introduced in "dma-buf: dma-heap: don't publish fd before copy_to_user() succeeds" [1].

[1] https://lore.kernel.org/dri-devel/[email protected]/

---
 drivers/gpu/drm/drm_prime.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 9b44c78cd77f..fe3436d1235d 100644
@@ -524,7 +524,7 @@ int drm_gem_prime_handle_to_fd(struct drm_device *dev,
 		return PTR_ERR(dmabuf);
 	}
 
-	fd_install(fd, dmabuf->file);
+	dma_buf_fd_install(dmabuf, fd);
 	*prime_fd = fd;
 	return 0;
 }

2.34.1