Skip to contentSkip to conversation

net: macb: fix CONFIG_OF=n build

PATCH In review Zhixing Chen 3 participants 4 messages Aug 19 – Aug 20

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: macb: fix CONFIG_OF=n build

macb_alloc_tieoff() and macb_free_tieoff() are called from macb_probe() and macb_remove(), which are built regardless of CONFIG_OF.

The helpers were added inside the CONFIG_OF block, although the tieoff descriptor handling does not depend on OF. This breaks CONFIG_OF=n builds with implicit function declaration errors.

Move the helpers outside the CONFIG_OF block.

Fixes 5262eab9462a ("net: macb: allocate tieoff descriptor once across device lifetime")

---

I hit this while building the latest net-next tree with my local config, where CONFIG_MACB is enabled and CONFIG_OF is disabled.

Since CONFIG_MACB does not depend on CONFIG_OF, macb_main.c still needs to build in this configuration. However, macb_probe() and macb_remove() are built while macb_alloc_tieoff() and macb_free_tieoff() are hidden behind CONFIG_OF.

Build error:

drivers/net/ethernet/cadence/macb_main.c: In function 'macb_probe':
drivers/net/ethernet/cadence/macb_main.c:5951:15: error: implicit declaration of function 'macb_alloc_tieoff' [-Werror=implicit-function-declaration]
5951 | err = macb_alloc_tieoff(bp);
drivers/net/ethernet/cadence/macb_main.c:5973:9: error: implicit declaration of function 'macb_free_tieoff' [-Werror=implicit-function-declaration]
5973 | macb_free_tieoff(bp);
CC [M] drivers/gpu/drm/i915/display/intel_qp_tables.o

---
 drivers/net/ethernet/cadence/macb_main.c | 64 ++++++++++++------------
 1 file changed, 32 insertions(+), 32 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +32 -32
index 1476bce77f34..c7e3b630b49e 100644
@@ -5507,38 +5507,6 @@ static int eyeq5_init(struct platform_device *pdev)
 	return ret;
 }
 
-static int macb_alloc_tieoff(struct macb *bp)
-{
-	/* Tieoff is a workaround in case HW cannot disable queues, for PM. */
-	if (bp->caps & MACB_CAPS_QUEUE_DISABLE)
-		return 0;
-
-	bp->rx_ring_tieoff = dma_alloc_coherent(&bp->pdev->dev,
-						macb_dma_desc_get_size(bp),
-						&bp->rx_ring_tieoff_dma,
-						GFP_KERNEL);
-	if (!bp->rx_ring_tieoff)
-		return -ENOMEM;
-
-	macb_set_addr(bp, bp->rx_ring_tieoff,
-		      MACB_BIT(RX_WRAP) | MACB_BIT(RX_USED));
-
-	bp->rx_ring_tieoff->ctrl = 0;
-
-	return 0;
-}
-
-static void macb_free_tieoff(struct macb *bp)
-{
-	if (!bp->rx_ring_tieoff)
-		return;
-
-	dma_free_coherent(&bp->pdev->dev, macb_dma_desc_get_size(bp),
-			  bp->rx_ring_tieoff,
-			  bp->rx_ring_tieoff_dma);
-	bp->rx_ring_tieoff = NULL;
-}
-
 static const struct macb_usrio_config mpfs_usrio = {
 	.tsu_source = 0,
 };
@@ -5756,6 +5724,38 @@ static const struct of_device_id macb_dt_ids[] = {
 MODULE_DEVICE_TABLE(of, macb_dt_ids);
 #endif /* CONFIG_OF */
 
+static int macb_alloc_tieoff(struct macb *bp)
+{
+	/* Tieoff is a workaround in case HW cannot disable queues, for PM. */
+	if (bp->caps & MACB_CAPS_QUEUE_DISABLE)
+		return 0;
+
+	bp->rx_ring_tieoff = dma_alloc_coherent(&bp->pdev->dev,
+						macb_dma_desc_get_size(bp),
+						&bp->rx_ring_tieoff_dma,
+						GFP_KERNEL);
+	if (!bp->rx_ring_tieoff)
+		return -ENOMEM;
+
+	macb_set_addr(bp, bp->rx_ring_tieoff,
+		      MACB_BIT(RX_WRAP) | MACB_BIT(RX_USED));
+
+	bp->rx_ring_tieoff->ctrl = 0;
+
+	return 0;
+}
+
+static void macb_free_tieoff(struct macb *bp)
+{
+	if (!bp->rx_ring_tieoff)
+		return;
+
+	dma_free_coherent(&bp->pdev->dev, macb_dma_desc_get_size(bp),
+			  bp->rx_ring_tieoff,
+			  bp->rx_ring_tieoff_dma);
+	bp->rx_ring_tieoff = NULL;
+}
+
 static const struct macb_config default_gem_config = {
 	.caps = MACB_CAPS_GIGABIT_MODE_AVAILABLE |
 		MACB_CAPS_JUMBO |

base-commit: 61eb236c41c2a4717015dff18016a75a5eb90052

Signature · 1 line
2.34.1
#1 lore raw

Hi Zhixing

Quoted 23 lines · Zhixing Chen

macb_alloc_tieoff() and macb_free_tieoff() are called from macb_probe() and macb_remove(), which are built regardless of CONFIG_OF.

The helpers were added inside the CONFIG_OF block, although the tieoff descriptor handling does not depend on OF. This breaks CONFIG_OF=n builds with implicit function declaration errors.

Move the helpers outside the CONFIG_OF block.

Fixes 5262eab9462a ("net: macb: allocate tieoff descriptor once across device lifetime")

---

I hit this while building the latest net-next tree with my local config, where CONFIG_MACB is enabled and CONFIG_OF is disabled.

Since CONFIG_MACB does not depend on CONFIG_OF, macb_main.c still needs to build in this configuration. However, macb_probe() and macb_remove() are built while macb_alloc_tieoff() and macb_free_tieoff() are hidden behind CONFIG_OF.

Build error:

Quoted 8 lines

drivers/net/ethernet/cadence/macb_main.c: In function 'macb_probe':
drivers/net/ethernet/cadence/macb_main.c:5951:15: error: implicit declaration of function 'macb_alloc_tieoff' [-Werror=implicit-function-declaration]
5951 | err = macb_alloc_tieoff(bp);
drivers/net/ethernet/cadence/macb_main.c:5973:9: error: implicit declaration of function 'macb_free_tieoff' [-Werror=implicit-function-declaration]
5973 | macb_free_tieoff(bp);
CC [M] drivers/gpu/drm/i915/display/intel_qp_tables.o

---

Nathan sent the same patch a few hours earlier [1]. I will follow up there.

Quoted 1 line

[...]

[1] https://lore.kernel.org/netdev/[email protected]/

Thanks,
Nicolai

#2 lore raw
Quoted 9 lines

Hi Zhixing

Quoted 1 line · Zhixing Chen

macb_alloc_tieoff() and macb_free_tieoff() are called from macb_probe()

Nathan sent the same patch a few hours earlier [1]. I will follow up there.

Quoted 7 lines
Quoted 1 line

[...]

[1]
https://lore.kernel.org/netdev/[email protected]/

Thanks,
Nicolai

Hi Nicolai,

Thanks for the pointer.

I missed Nathan's earlier patch before sending mine. Please disregard my patch.

Thanks,
Zhixing

maintainer
#3 lore raw

Hello Zhixing,

Quoted 10 lines · Zhixing Chen
Quoted 5 lines · Zhixing Chen
Quoted 1 line

macb_alloc_tieoff() and macb_free_tieoff() are called from macb_probe()

Nathan sent the same patch a few hours earlier [1]. I will follow up there.

Thanks for the pointer.

I missed Nathan's earlier patch before sending mine. Please disregard my patch.

I sent a patch which I prefer over the move-functions-around proposal. Goal being to avoid reproducing the same issue a second time. I forgot CCing, sorry.

https://lore.kernel.org/netdev/[email protected]/

Thanks,

--
Théo Lebrun, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com