drivers: pci: hailo: Fix kernel warning when calling find_vdma()

[naushir]

Calling this function without holding the mmap_read_lock causes the kernel to throw an error message, spamming the dmesg logs when running the Hailo hardware.

Fix it by adding the approprite lock/unlock functions around find_vdma().

[naushir]

The formatting for this patch is all messed and will fail the check, but matches the original source code from Hailo.

[naushir]

Let's hold off from merging this until Hailo have a change to review - I've pinged them already.

[tsailerddln]

Unfortunately, while this fix silences the dmesg spamming, but it causes the hailo driver to hang after a minute or ten.

[HailoRT] [error] CHECK failed - Waiting for async job to finish has failed with timeout (1000ms)
[HailoRT] [error] CHECK_SUCCESS failed with status=HAILO_TIMEOUT(4)

And the kernel:
[ 1209.578099] Call trace:
[ 1209.578101] __switch_to+0xf0/0x160
[ 1209.578111] __schedule+0x32c/0xb60
[ 1209.578115] schedule+0x3c/0x148
[ 1209.578119] schedule_preempt_disabled+0x2c/0x50
[ 1209.578124] rwsem_down_read_slowpath+0x1e0/0x510
[ 1209.578127] down_read+0xb0/0xc0
[ 1209.578130] hailo_vdma_buffer_map+0x3c8/0x6b0 [hailo_pci]
[ 1209.578141] hailo_vdma_buffer_map_ioctl+0xcc/0x348 [hailo_pci]
[ 1209.578146] hailo_vdma_ioctl+0x254/0x2a8 [hailo_pci]
[ 1209.578151] hailo_pcie_fops_unlockedioctl+0x16c/0x7e8 [hailo_pci]
[ 1209.578157] __arm64_sys_ioctl+0xb0/0x100
[ 1209.578162] invoke_syscall+0x50/0x120
[ 1209.578167] el0_svc_common.constprop.0+0x48/0xf8
[ 1209.578171] do_el0_svc+0x28/0x40
[ 1209.578175] el0_svc+0x30/0xd0
[ 1209.578179] el0t_64_sync_handler+0x13c/0x158
[ 1209.578182] el0t_64_sync+0x190/0x198

[naushir]

@tsailerddln what are you running to cause this error? I'll try to reproduce this on my end.

EDIT: It might be useful to try out the change in #6868 to see if it helps improve your stability.