diff mbox series

[net,v2,2/2] net/txgbe: add extra handle for MSI/INTx into thread irq handle

Message ID 20240626060703.31652-3-jiawenwu@trustnetic.com (mailing list archive)
State Superseded
Delegated to: Netdev Maintainers
Headers show
Series net: txgbe: fix MSI and INTx interrupts | expand

Checks

Context Check Description
netdev/series_format success Posting correctly formatted
netdev/tree_selection success Clearly marked for net
netdev/ynl success Generated files up to date; no warnings/errors; no diff in generated;
netdev/fixes_present success Fixes tag present in non-next series
netdev/header_inline success No static functions without inline keyword in header files
netdev/build_32bit success Errors and warnings before: 859 this patch: 859
netdev/build_tools success No tools touched, skip
netdev/cc_maintainers success CCed 7 of 7 maintainers
netdev/build_clang success Errors and warnings before: 863 this patch: 863
netdev/verify_signedoff success Signed-off-by tag matches author and committer
netdev/deprecated_api success None detected
netdev/check_selftest success No net selftest shell script
netdev/verify_fixes success Fixes tag looks correct
netdev/build_allmodconfig_warn success Errors and warnings before: 863 this patch: 863
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 106 lines checked
netdev/build_clang_rust success No Rust files in patch. Skipping build
netdev/kdoc success Errors and warnings before: 21 this patch: 21
netdev/source_inline success Was 0 now: 0
netdev/contest success net-next-2024-06-27--03-00 (tests: 665)

Commit Message

Jiawen Wu June 26, 2024, 6:07 a.m. UTC 
Rename original txgbe_misc_irq_handle() to txgbe_misc_irq_thread_fn()
since it is the handle thread to wake up. And add the primary handler
to deal the case of MSI/INTx, because there is a schedule NAPI poll.

Moreover, do not free isb resources in .ndo_stop, to avoid reading
memory by a null pointer.

Fixes: aefd013624a1 ("net: txgbe: use irq_domain for interrupt controller")
Signed-off-by: Jiawen Wu <jiawenwu@trustnetic.com>
---
 drivers/net/ethernet/wangxun/libwx/wx_lib.c   | 10 ++++-
 .../net/ethernet/wangxun/txgbe/txgbe_irq.c    | 44 ++++++++++++++++---
 .../net/ethernet/wangxun/txgbe/txgbe_main.c   |  1 +
 3 files changed, 49 insertions(+), 6 deletions(-)

Comments

Jakub Kicinski June 27, 2024, 11:43 p.m. UTC | #1 
On Wed, 26 Jun 2024 14:07:03 +0800 Jiawen Wu wrote:
> Moreover, do not free isb resources in .ndo_stop, to avoid reading
> memory by a null pointer.

Please provide more detail on the sequence of events leading to the
null-defer.

>  	pdev->irq = pci_irq_vector(pdev, 0);
> +	wx->num_q_vectors = 1;

this doesn't seem obviously related

>  
>  	return 0;
>  }
> @@ -2027,6 +2028,9 @@ int wx_setup_isb_resources(struct wx *wx)
>  {
>  	struct pci_dev *pdev = wx->pdev;
>  
> +	if (wx->isb_mem)
> +		return 0;
> +
>  	wx->isb_mem = dma_alloc_coherent(&pdev->dev,
>  					 sizeof(u32) * 4,
>  					 &wx->isb_dma,
> @@ -2050,6 +2054,9 @@ void wx_free_isb_resources(struct wx *wx)
>  {
>  	struct pci_dev *pdev = wx->pdev;
>  
> +	if (!wx->isb_mem)
> +		return;
> +

And neither does this. Why do you need to make these function
idempotent?
Jiawen Wu June 28, 2024, 2:16 a.m. UTC | #2 
On Fri, Jun 28, 2024 7:44 AM, Jakub Kicinski wrote:
> On Wed, 26 Jun 2024 14:07:03 +0800 Jiawen Wu wrote:
> > Moreover, do not free isb resources in .ndo_stop, to avoid reading
> > memory by a null pointer.
> 
> Please provide more detail on the sequence of events leading to the
> null-defer.
> 
> >  	pdev->irq = pci_irq_vector(pdev, 0);
> > +	wx->num_q_vectors = 1;
> 
> this doesn't seem obviously related

Umm, this is related another fix for MSI/INTx. I'll split it into a separate
commit.

> 
> >
> >  	return 0;
> >  }
> > @@ -2027,6 +2028,9 @@ int wx_setup_isb_resources(struct wx *wx)
> >  {
> >  	struct pci_dev *pdev = wx->pdev;
> >
> > +	if (wx->isb_mem)
> > +		return 0;
> > +
> >  	wx->isb_mem = dma_alloc_coherent(&pdev->dev,
> >  					 sizeof(u32) * 4,
> >  					 &wx->isb_dma,
> > @@ -2050,6 +2054,9 @@ void wx_free_isb_resources(struct wx *wx)
> >  {
> >  	struct pci_dev *pdev = wx->pdev;
> >
> > +	if (!wx->isb_mem)
> > +		return;
> > +
> 
> And neither does this. Why do you need to make these function
> idempotent?

This is also to implement txgbe changes without changing the flow of ngbe.
diff mbox series

Patch

diff --git a/drivers/net/ethernet/wangxun/libwx/wx_lib.c b/drivers/net/ethernet/wangxun/libwx/wx_lib.c
index 99f55a3573c8..f098758893b3 100644
--- a/drivers/net/ethernet/wangxun/libwx/wx_lib.c
+++ b/drivers/net/ethernet/wangxun/libwx/wx_lib.c
@@ -1686,6 +1686,7 @@  static int wx_set_interrupt_capability(struct wx *wx)
 	}
 
 	pdev->irq = pci_irq_vector(pdev, 0);
+	wx->num_q_vectors = 1;
 
 	return 0;
 }
@@ -2027,6 +2028,9 @@  int wx_setup_isb_resources(struct wx *wx)
 {
 	struct pci_dev *pdev = wx->pdev;
 
+	if (wx->isb_mem)
+		return 0;
+
 	wx->isb_mem = dma_alloc_coherent(&pdev->dev,
 					 sizeof(u32) * 4,
 					 &wx->isb_dma,
@@ -2050,6 +2054,9 @@  void wx_free_isb_resources(struct wx *wx)
 {
 	struct pci_dev *pdev = wx->pdev;
 
+	if (!wx->isb_mem)
+		return;
+
 	dma_free_coherent(&pdev->dev, sizeof(u32) * 4,
 			  wx->isb_mem, wx->isb_dma);
 	wx->isb_mem = NULL;
@@ -2386,7 +2393,8 @@  static void wx_free_all_tx_resources(struct wx *wx)
 
 void wx_free_resources(struct wx *wx)
 {
-	wx_free_isb_resources(wx);
+	if (wx->mac.type == wx_mac_em)
+		wx_free_isb_resources(wx);
 	wx_free_all_rx_resources(wx);
 	wx_free_all_tx_resources(wx);
 }
diff --git a/drivers/net/ethernet/wangxun/txgbe/txgbe_irq.c b/drivers/net/ethernet/wangxun/txgbe/txgbe_irq.c
index ac789ec0091a..15e0fef02aac 100644
--- a/drivers/net/ethernet/wangxun/txgbe/txgbe_irq.c
+++ b/drivers/net/ethernet/wangxun/txgbe/txgbe_irq.c
@@ -111,6 +111,36 @@  static const struct irq_domain_ops txgbe_misc_irq_domain_ops = {
 };
 
 static irqreturn_t txgbe_misc_irq_handle(int irq, void *data)
+{
+	struct wx_q_vector *q_vector;
+	struct txgbe *txgbe = data;
+	struct wx *wx = txgbe->wx;
+	u32 eicr;
+
+	if (wx->pdev->msix_enabled)
+		return IRQ_WAKE_THREAD;
+
+	eicr = wx_misc_isb(wx, WX_ISB_VEC0);
+	if (!eicr) {
+		/* shared interrupt alert!
+		 * the interrupt that we masked before the ICR read.
+		 */
+		if (netif_running(wx->netdev))
+			txgbe_irq_enable(wx, true);
+		return IRQ_NONE;        /* Not our interrupt */
+	}
+	wx->isb_mem[WX_ISB_VEC0] = 0;
+	if (!(wx->pdev->msi_enabled))
+		wr32(wx, WX_PX_INTA, 1);
+
+	/* would disable interrupts here but it is auto disabled */
+	q_vector = wx->q_vector[0];
+	napi_schedule_irqoff(&q_vector->napi);
+
+	return IRQ_WAKE_THREAD;
+}
+
+static irqreturn_t txgbe_misc_irq_thread_fn(int irq, void *data)
 {
 	struct txgbe *txgbe = data;
 	struct wx *wx = txgbe->wx;
@@ -157,6 +187,7 @@  void txgbe_free_misc_irq(struct txgbe *txgbe)
 
 int txgbe_setup_misc_irq(struct txgbe *txgbe)
 {
+	unsigned long flags = IRQF_ONESHOT;
 	struct wx *wx = txgbe->wx;
 	int hwirq, err;
 
@@ -170,14 +201,17 @@  int txgbe_setup_misc_irq(struct txgbe *txgbe)
 		irq_create_mapping(txgbe->misc.domain, hwirq);
 
 	txgbe->misc.chip = txgbe_irq_chip;
-	if (wx->pdev->msix_enabled)
+	if (wx->pdev->msix_enabled) {
 		txgbe->misc.irq = wx->msix_entry->vector;
-	else
+	} else {
 		txgbe->misc.irq = wx->pdev->irq;
+		if (!wx->pdev->msi_enabled)
+			flags |= IRQF_SHARED;
+	}
 
-	err = request_threaded_irq(txgbe->misc.irq, NULL,
-				   txgbe_misc_irq_handle,
-				   IRQF_ONESHOT,
+	err = request_threaded_irq(txgbe->misc.irq, txgbe_misc_irq_handle,
+				   txgbe_misc_irq_thread_fn,
+				   flags,
 				   wx->netdev->name, txgbe);
 	if (err)
 		goto del_misc_irq;
diff --git a/drivers/net/ethernet/wangxun/txgbe/txgbe_main.c b/drivers/net/ethernet/wangxun/txgbe/txgbe_main.c
index 76b5672c0a17..92c1fae826d0 100644
--- a/drivers/net/ethernet/wangxun/txgbe/txgbe_main.c
+++ b/drivers/net/ethernet/wangxun/txgbe/txgbe_main.c
@@ -729,6 +729,7 @@  static void txgbe_remove(struct pci_dev *pdev)
 
 	txgbe_remove_phy(txgbe);
 	txgbe_free_misc_irq(txgbe);
+	wx_free_isb_resources(wx);
 
 	pci_release_selected_regions(pdev,
 				     pci_select_bars(pdev, IORESOURCE_MEM));