3.16.50-rc1 review patch. If anyone has any objections, please let me know.
------------------
From: Shu Wang <shuwang@xxxxxxxxxx>
commit d6f5f071f1e13cadecf8aef1faa7e5d6fbc9f33b upstream.
Found this issue by kmemleak.
xhci_run() did not check return val and free command for
xhci_queue_vendor_command()
unreferenced object 0xffff88011c0be500 (size 64):
comm "kworker/0:1", pid 58, jiffies 4294670908 (age 50.420s)
hex dump (first 32 bytes):
backtrace:
[<ffffffff8176166a>] kmemleak_alloc+0x4a/0xa0
[<ffffffff8121801a>] kmem_cache_alloc_trace+0xca/0x1d0
[<ffffffff81576bf4>] xhci_alloc_command+0x44/0x130
[<ffffffff8156f1cc>] xhci_run+0x4cc/0x630
[<ffffffff8153b84b>] usb_add_hcd+0x3bb/0x950
[<ffffffff8154eac8>] usb_hcd_pci_probe+0x188/0x500
[<ffffffff815851ac>] xhci_pci_probe+0x2c/0x220
[<ffffffff813d2ca5>] local_pci_probe+0x45/0xa0
[<ffffffff810a54e4>] work_for_cpu_fn+0x14/0x20
[<ffffffff810a8409>] process_one_work+0x149/0x360
[<ffffffff810a8d08>] worker_thread+0x1d8/0x3c0
[<ffffffff810ae7d9>] kthread+0x109/0x140
[<ffffffff8176d585>] ret_from_fork+0x25/0x30
[<ffffffffffffffff>] 0xffffffffffffffff
Signed-off-by: Shu Wang <shuwang@xxxxxxxxxx>
Signed-off-by: Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
---
drivers/usb/host/xhci.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
--- a/drivers/usb/host/xhci.c
+++ b/drivers/usb/host/xhci.c
@@ -662,8 +662,10 @@ int xhci_run(struct usb_hcd *hcd)
command = xhci_alloc_command(xhci, false, false, GFP_KERNEL);
if (!command)
return -ENOMEM;
- xhci_queue_vendor_command(xhci, command, 0, 0, 0,
+ ret = xhci_queue_vendor_command(xhci, command, 0, 0, 0,
TRB_TYPE(TRB_NEC_GET_FW));
+ if (ret)
+ xhci_free_command(xhci, command);
}
xhci_dbg_trace(xhci, trace_xhci_dbg_init,
"Finished xhci_run for USB2 roothub");