[PATCH v2 0/4] treewide: fixup gfp_t printks

[Brendan Jackman]

This patchset used to be about switching gfp_t to unsigned long. That is
probably not gonna happen any more but while writing it I found these
cleanups that seem worthwhile regardless.

Signed-off-by: Brendan Jackman <jack...@google.com>
---
Changes in v2:
- Drop gfp_t changes
- Add correct CCs
- Add minor fixups to preexisting code spotted by AI review
- Link to v1: https://lore.kernel.org/r/20260319-gfp64-v...@google.com

---
Brendan Jackman (4):
drm/managed: Use special gfp_t format specifier
iwlegacy: 3945-mac: Fixup allocation failure log
mm/kfence: Use special gfp_t format specifier
net/rds: Use special gfp_t format specifier

drivers/gpu/drm/drm_managed.c | 4 ++--
drivers/net/wireless/intel/iwlegacy/3945-mac.c | 4 ++--
mm/kfence/kfence_test.c | 2 +-
net/rds/tcp_recv.c | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
---
base-commit: c369299895a591d96745d6492d4888259b004a9e
change-id: 20260319-gfp64-7a970a80ba4e

Best regards,
--
Brendan Jackman <jack...@google.com>

[Brendan Jackman]

%pGg produces nice readable output and decouples the format string from
the size of gfp_t.

Signed-off-by: Brendan Jackman <jack...@google.com>
---

drivers/gpu/drm/drm_managed.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/drm_managed.c b/drivers/gpu/drm/drm_managed.c
index 247f468731de0..a9da94319b052 100644
--- a/drivers/gpu/drm/drm_managed.c
+++ b/drivers/gpu/drm/drm_managed.c
@@ -232,8 +232,8 @@ void *drmm_kmalloc(struct drm_device *dev, size_t size, gfp_t gfp)

dr = alloc_dr(NULL, size, gfp, dev_to_node(dev->dev));
if (!dr) {
- drm_dbg_drmres(dev, "failed to allocate %zu bytes, %u flags\n",
- size, gfp);
+ drm_dbg_drmres(dev, "failed to allocate %zu bytes, %pGg\n",
+ size, &gfp);
return NULL;
}
dr->node.name = kstrdup_const("kmalloc", gfp);

--
2.51.2

[Brendan Jackman]

Fix 2 issues spotted by AI[0]:

1. Missing space after the full stop.

2. Wrong GFP flags are printed.

And also switch to %pGg for the GFP flags. This produces nice readable

output and decouples the format string from the size of gfp_t.

[0] https://sashiko.dev/#/patchset/20260319-gfp64-v1-0-2c73b8d42b7f%40google.com

Signed-off-by: Brendan Jackman <jack...@google.com>
---

drivers/net/wireless/intel/iwlegacy/3945-mac.c | 4 ++--

1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/wireless/intel/iwlegacy/3945-mac.c b/drivers/net/wireless/intel/iwlegacy/3945-mac.c
index c148654aa9533..88b31e0b9568c 100644
--- a/drivers/net/wireless/intel/iwlegacy/3945-mac.c
+++ b/drivers/net/wireless/intel/iwlegacy/3945-mac.c
@@ -1002,9 +1002,9 @@ il3945_rx_allocate(struct il_priv *il, gfp_t priority)
D_INFO("Failed to allocate SKB buffer.\n");
if (rxq->free_count <= RX_LOW_WATERMARK &&
net_ratelimit())
- IL_ERR("Failed to allocate SKB buffer with %0x."
+ IL_ERR("Failed to allocate SKB buffer with %pGg. "
"Only %u free buffers remaining.\n",
- priority, rxq->free_count);
+ &gfp_mask, rxq->free_count);
/* We don't reschedule replenish work here -- we will
* call the restock method and if it still needs
* more buffers it will schedule replenish */

--
2.51.2

[Brendan Jackman]

%pGg produces nice readable output and decouples the format string from
the size of gfp_t.

Signed-off-by: Brendan Jackman <jack...@google.com>
---

mm/kfence/kfence_test.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/kfence/kfence_test.c b/mm/kfence/kfence_test.c
index 5725a367246d9..10424cd25e5a6 100644
--- a/mm/kfence/kfence_test.c
+++ b/mm/kfence/kfence_test.c
@@ -263,7 +263,7 @@ static void *test_alloc(struct kunit *test, size_t size, gfp_t gfp, enum allocat
break;
}

- kunit_info(test, "%s: size=%zu, gfp=%x, policy=%s, cache=%i\n", __func__, size, gfp,
+ kunit_info(test, "%s: size=%zu, gfp=%pGg, policy=%s, cache=%i\n", __func__, size, &gfp,
policy_name, !!test_cache);

/*

--
2.51.2

[Brendan Jackman]

%pGg produces nice readable output and decouples the format string from
the size of gfp_t.

Signed-off-by: Brendan Jackman <jack...@google.com>
---

net/rds/tcp_recv.c | 2 +-

1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/rds/tcp_recv.c b/net/rds/tcp_recv.c
index 49f96ee0c40f6..ffe843ca219c7 100644
--- a/net/rds/tcp_recv.c
+++ b/net/rds/tcp_recv.c
@@ -275,7 +275,7 @@ static int rds_tcp_read_sock(struct rds_conn_path *cp, gfp_t gfp)
desc.count = 1; /* give more than one skb per call */

tcp_read_sock(sock->sk, &desc, rds_tcp_data_recv);
- rdsdebug("tcp_read_sock for tc %p gfp 0x%x returned %d\n", tc, gfp,
+ rdsdebug("tcp_read_sock for tc %p gfp %pGg returned %d\n", tc, &gfp,
desc.error);

if (skb_queue_empty_lockless(&sock->sk->sk_receive_queue) &&

--
2.51.2

[Stanislaw Gruszka]

Hi,

On Thu, Mar 26, 2026 at 12:31:58PM +0000, Brendan Jackman wrote:
> Fix 2 issues spotted by AI[0]:
>
> 1. Missing space after the full stop.
>
> 2. Wrong GFP flags are printed.

We should also initialize gfp_mask = priority inside the loop.
But this can be done as separate patch.

> And also switch to %pGg for the GFP flags. This produces nice readable
> output and decouples the format string from the size of gfp_t.
>
> [0] https://sashiko.dev/#/patchset/20260319-gfp64-v1-0-2c73b8d42b7f%40google.com
>
> Signed-off-by: Brendan Jackman <jack...@google.com>

Acked-by: Stanislaw Gruszka <stf...@wp.pl>

[Stanislaw Gruszka]

On Thu, Mar 26, 2026 at 12:31:56PM +0000, Brendan Jackman wrote:
> This patchset used to be about switching gfp_t to unsigned long. That is
> probably not gonna happen any more but while writing it I found these
> cleanups that seem worthwhile regardless.

For the record, these patches are independent and could be sent separately
to the appropriate maintainers. Doing so would avoid unnecessarily large CC
lists for people not interested about the changes in other subsystems.

Regards
Stanislaw

[Allison Henderson]

This looks fine to me. Thanks Brendan!

Acked-by: Allison Henderson <ache...@kernel.org>