From 17451138cf60f5d64eed88cc5defd44981926d9d Mon Sep 17 00:00:00 2001 From: Konsta Holtta Date: Thu, 31 Aug 2017 13:01:26 +0300 Subject: gpu: nvgpu: hold ch ref when getting ch from fd Fix a race condition in gk20a_get_channel_from_file() that returns a channel pointer from an fd: take a reference to the channel before putting the file ref back. Now the caller is responsible of releasing the channel reference eventually. Also document why dbg_session_channel_data has to hold a ref to the channel file instead of just the channel: that might deadlock if the fds were closed in "wrong" order. Change-Id: I8e91b809f5f7b1cb0c1487bd955ad6d643727a53 Signed-off-by: Konsta Holtta Reviewed-on: https://git-master.nvidia.com/r/1549290 Reviewed-by: mobile promotions Tested-by: mobile promotions --- drivers/gpu/nvgpu/common/linux/ioctl_tsg.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/gpu/nvgpu/common/linux/ioctl_tsg.c') diff --git a/drivers/gpu/nvgpu/common/linux/ioctl_tsg.c b/drivers/gpu/nvgpu/common/linux/ioctl_tsg.c index c68c907e..d35ea14c 100644 --- a/drivers/gpu/nvgpu/common/linux/ioctl_tsg.c +++ b/drivers/gpu/nvgpu/common/linux/ioctl_tsg.c @@ -49,6 +49,8 @@ static int gk20a_tsg_bind_channel_fd(struct tsg_gk20a *tsg, int ch_fd) return -EINVAL; err = ch->g->ops.fifo.tsg_bind_channel(tsg, ch); + + gk20a_channel_put(ch); return err; } -- cgit v1.2.2