From: Arseny Solokha <asolo...@kb.kras.ru>
When DMA-API debugging is enabled in the kernel, it spews the following
upon upping the link:
fsl-gianfar ffe25000.ethernet: DMA-API: device driver failed to check map
error[device address=0x0000000005f41012] [size=90 bytes] [map-
WARNING: at lib/dma-debug.c:1135
Modules linked in:
CPU: 1 PID: 0 Comm: swapper/1 Tainted: G O 3.18.0-rc7 #1
task: ee06f080 ti: effde000 task.ti: ee0b2000
NIP: c01d7c1c LR: c01d7c1c CTR: 00000000
REGS: effdfd40 TRAP: 0700 Tainted: G O (3.18.0-rc7)
MSR: 00021000 <CE,ME> CR: 42804442 XER: 00000000
GPR00: c01d7c1c effdfdf0 ee06f080 00000097 00000001 c0066dd8 00000000 00000001
GPR08: 00000000 00000000 effde000 0000029e 00000000 00000000 c55fa740 ee0d6818
GPR16: r00000600 c5a6a9c0 ee0d6830 ee0d6850 ffff8100 00000008 00000000 c5bbc800
GPR24: c0730000 00029000 c0d0b828 c072c394 effdfe48 c075baec c0d0f020 ee308300
NIP [c01d7c1c] check_unmap+0x5b4/0xae4
LR [c01d7c1c] check_unmap+0x5b4/0xae4
Call Trace:
[effdfdf0] [c01d7c1c] check_unmap+0x5b4/0xae4 (unreliable)
[effdfe40] [c01d81c4] debug_dma_unmap_page+0x78/0x8c
[effdfec0] [c028b270] gfar_clean_rx_ring+0x114/0x4c0
[effdff30] [c028b814] gfar_poll_rx_sq+0x3c/0xa4
[effdff50] [c030c388] net_rx_action+0x130/0x1ac
[effdff80] [c00319e0] __do_softirq+0x134/0x240
[effdffe0] [c0031dd0] irq_exit+0xa4/0xc8
[effdfff0] [c000e01c] call_do_irq+0x24/0x3c
[ee0b3e60] [c0004a04] do_IRQ+0x8c/0x108
[ee0b3e80] [c0010068] ret_from_except+0x0/0x18
--- interrupt: 501 at arch_cpu_idle+0x24/0x5c
LR = arch_cpu_idle+0x24/0x5c
[ee0b3f40] [c007d2e4] rcu_idle_enter+0xc8/0xcc (unreliable)
[ee0b3f50] [c006587c] cpu_startup_entry+0x1d4/0x29c
[ee0b3fa0] [c00111cc] start_secondary+0x364/0x478
[ee0b3ff0] [c000217c] __secondary_start+0x7c/0xc8
Instruction dump:
394adb30 80fc0018 811c001c 3c60c04f 5529103a 7cca482e 38639e60 813c0020
815c0024 90c10008 4cc63182 48240b01 <0fe00000> 3c60c04f 3863971c 4cc63182
---[ end trace 3eb7bf62ba1b80f8 ]---
oMapped at:
[<c02887cc>] gfar_new_rxbdp.isra.4+0x120/0x16c
[<c0288968>] gfar_init_bds+0x150/0x1b0
[<c028a800>] startup_gfar+0x334/0x3d8
[<c028ac64>] gfar_enet_open+0x2b8/0x460
[<c03100c0>] __dev_open+0xdc/0x150
And the underlying code indeed doesn't perform the check.
Signed-off-by: Arseny Solokha <asolo...@kb.kras.ru>
---
drivers/net/ethernet/freescale/gianfar.c | 32 ++++++++++++++++++++++++++------
1 file changed, 26 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/freescale/gianfar.c
b/drivers/net/ethernet/freescale/gianfar.c
index 4fdf0aa..f34ca55 100644
--- a/drivers/net/ethernet/freescale/gianfar.c
+++ b/drivers/net/ethernet/freescale/gianfar.c
@@ -117,8 +117,8 @@ static void gfar_reset_task(struct work_struct *work);
static void gfar_timeout(struct net_device *dev);
static int gfar_close(struct net_device *dev);
struct sk_buff *gfar_new_skb(struct net_device *dev);
-static void gfar_new_rxbdp(struct gfar_priv_rx_q *rx_queue, struct rxbd8 *bdp,
- struct sk_buff *skb);
+static int gfar_new_rxbdp(struct gfar_priv_rx_q *rx_queue, struct rxbd8 *bdp,
+ struct sk_buff *skb);
static int gfar_set_mac_address(struct net_device *dev);
static int gfar_change_mtu(struct net_device *dev, int new_mtu);
static irqreturn_t gfar_error(int irq, void *dev_id);
@@ -214,6 +214,8 @@ static int gfar_init_bds(struct net_device *ndev)
gfar_init_rxbdp(rx_queue, rxbdp,
rxbdp->bufPtr);
} else {
+ int ret;
+
skb = gfar_new_skb(ndev);
if (!skb) {
netdev_err(ndev, "Can't allocate RX
buffers\n");
@@ -221,7 +223,11 @@ static int gfar_init_bds(struct net_device *ndev)
}
rx_queue->rx_skbuff[j] = skb;
- gfar_new_rxbdp(rx_queue, rxbdp, skb);
+ ret = gfar_new_rxbdp(rx_queue, rxbdp, skb);
+ if (ret) {
+ netdev_err(ndev, "Buffer mapping
error\n");
+ return ret;
+ }
}
rxbdp++;
@@ -2606,8 +2612,8 @@ static void gfar_clean_tx_ring(struct gfar_priv_tx_q
*tx_queue)
netdev_tx_completed_queue(txq, howmany, bytes_sent);
}
-static void gfar_new_rxbdp(struct gfar_priv_rx_q *rx_queue, struct rxbd8 *bdp,
- struct sk_buff *skb)
+static int gfar_new_rxbdp(struct gfar_priv_rx_q *rx_queue, struct rxbd8 *bdp,
+ struct sk_buff *skb)
{
struct net_device *dev = rx_queue->dev;
struct gfar_private *priv = netdev_priv(dev);
@@ -2615,7 +2621,12 @@ static void gfar_new_rxbdp(struct gfar_priv_rx_q
*rx_queue, struct rxbd8 *bdp,
buf = dma_map_single(priv->dev, skb->data,
priv->rx_buffer_size, DMA_FROM_DEVICE);
+ if (dma_mapping_error(priv->dev, buf))
+ return -EFAULT;
+
gfar_init_rxbdp(rx_queue, bdp, buf);
+
+ return 0;
}
static struct sk_buff *gfar_alloc_skb(struct net_device *dev)
@@ -2805,6 +2816,7 @@ int gfar_clean_rx_ring(struct gfar_priv_rx_q *rx_queue,
int rx_work_limit)
while (!((bdp->status & RXBD_EMPTY) || (--rx_work_limit < 0))) {
struct sk_buff *newskb;
+ int rxbdpret;
rmb();
@@ -2854,7 +2866,15 @@ int gfar_clean_rx_ring(struct gfar_priv_rx_q *rx_queue,
int rx_work_limit)
rx_queue->rx_skbuff[rx_queue->skb_currx] = newskb;
/* Setup the new bdp */
- gfar_new_rxbdp(rx_queue, bdp, newskb);
+ rxbdpret = gfar_new_rxbdp(rx_queue, bdp, newskb);
+ if (unlikely(rxbdpret)) {
+ /* We drop the frame if we failed to map a new DMA
+ * buffer
+ */
+ count_errors(bdp->status, dev);
+ dev_kfree_skb(newskb);
+ continue;
+ }
/* Update to the next pointer */
bdp = next_bd(bdp, base, rx_queue->rx_ring_size);
--
2.2.0
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
