[PATCH 5.15 193/317] PCI/P2PDMA: Fix a sleeping issue in a RCU read section

[Sasha Levin]

From: Christophe JAILLET <christophe.jaillet@xxxxxxxxxx>

[ Upstream commit 1e5c66afd4a40bb7be17cb33cbb1a1085f727730 ]

It is not allowed to sleep within a RCU read section, so use GFP_ATOMIC
instead of GFP_KERNEL here.

Link: https://lore.kernel.org/r/02d9ec4a10235def0e764ff1f5be881ba12e16e8.1704397858.git.christophe.jaillet@xxxxxxxxxx
Fixes: ae21f835a5bd ("PCI/P2PDMA: Finish RCU conversion of pdev->p2pdma")
Signed-off-by: Christophe JAILLET <christophe.jaillet@xxxxxxxxxx>
Signed-off-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
Reviewed-by: Logan Gunthorpe <logang@xxxxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
 drivers/pci/p2pdma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/pci/p2pdma.c b/drivers/pci/p2pdma.c
index 316fd2f44df45..57654c82b08e8 100644
--- a/drivers/pci/p2pdma.c
+++ b/drivers/pci/p2pdma.c
@@ -536,7 +536,7 @@ calc_map_type_and_dist(struct pci_dev *provider, struct pci_dev *client,
 	p2pdma = rcu_dereference(provider->p2pdma);
 	if (p2pdma)
 		xa_store(&p2pdma->map_types, map_types_idx(client),
-			 xa_mk_value(map_type), GFP_KERNEL);
+			 xa_mk_value(map_type), GFP_ATOMIC);
 	rcu_read_unlock();
 	return map_type;
 }
-- 
2.43.0