In case p2p device was found but the p2p pool is empty, the nvme target
is still trying to free the sgl from the p2p pool instead of the
regular sgl pool and causing a crash (BUG() is called). Instead, assign
the p2p_dev for the request only if it was allocated from p2p pool.

This is the crash that was caused:

[Sun May 30 19:13:53 2021] ------------[ cut here ]------------
[Sun May 30 19:13:53 2021] kernel BUG at lib/genalloc.c:518!
[Sun May 30 19:13:53 2021] invalid opcode: 0000 [#1] SMP PTI
...
[Sun May 30 19:13:53 2021] kernel BUG at lib/genalloc.c:518!
...
[Sun May 30 19:13:53 2021] RIP: 0010:gen_pool_free_owner+0xa8/0xb0
...
[Sun May 30 19:13:53 2021] Call Trace:
[Sun May 30 19:13:53 2021] ------------[ cut here ]------------
[Sun May 30 19:13:53 2021]  pci_free_p2pmem+0x2b/0x70
[Sun May 30 19:13:53 2021]  pci_p2pmem_free_sgl+0x4f/0x80
[Sun May 30 19:13:53 2021]  nvmet_req_free_sgls+0x1e/0x80 [nvmet]
[Sun May 30 19:13:53 2021] kernel BUG at lib/genalloc.c:518!
[Sun May 30 19:13:53 2021]  nvmet_rdma_release_rsp+0x4e/0x1f0 [nvmet_rdma]
[Sun May 30 19:13:53 2021]  nvmet_rdma_send_done+0x1c/0x60 [nvmet_rdma]

Fixes: c6e3f133 ("nvmet: add metadata support for block devices")
Reviewed-by: Israel Rukshin <israelr@nvidia.com>
Signed-off-by: Max Gurtovoy <mgurtovoy@nvidia.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>