Skip to contentSkip to conversation

3c509.c: call SET_NETDEV_DEV for all device types (ISA/ISAPnP/EISA)

PATCH In review Matthew Whitehead 3 participants 7 messages Apr 25, 2013

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

PATCH 1/1 3c509.c: call SET_NETDEV_DEV for all device types (ISA/ISAPnP/EISA)

From: Matthew Whitehead <[email protected]>

The venerable 3c509 driver only sets its device parent in one case, the ISAPnP one. It does this with the SET_NETDEV_DEV function. It should register with the device hierarchy in two additional cases: standard (non-PnP) ISA and EISA.

- Currently they appear here:
/sys/devices/virtual/net/eth0 (standard ISA)
/sys/devices/virtual/net/eth1 (EISA)

- Rather, they should instead be here:
/sys/devices/isa/3c509.0/net/eth0 (standard ISA) /sys/devices/pci0000:00/0000:00:07.0/00:04/net/eth1 (EISA)

Tested on ISA and EISA boards.

---
 drivers/net/ethernet/3com/3c509.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)
drivers/net/ethernet/3com/3c509.c +2 -0
index f36ff99..adb4bf5 100644
@@ -306,6 +306,7 @@ static int el3_isa_match(struct device *pdev, unsigned int ndev)
 	if (!dev)
 		return -ENOMEM;
 
+	SET_NETDEV_DEV(dev, pdev);
 	netdev_boot_setup_check(dev);
 
 	if (!request_region(ioaddr, EL3_IO_EXTENT, "3c509-isa")) {
@@ -595,6 +596,7 @@ static int __init el3_eisa_probe (struct device *device)
 		return -ENOMEM;
 	}
 
+	SET_NETDEV_DEV(dev, device);
 	netdev_boot_setup_check(dev);
 
 	el3_dev_fill(dev, phys_addr, ioaddr, irq, if_port, EL3_EISA);
Signature · 1 line
1.7.2.5
#1 lore raw

Hello.

Quoted 10 lines · Matthew Whitehead

From: Matthew Whitehead <[email protected]>

The venerable 3c509 driver only sets its device parent in one case, the ISAPnP one. It does this with the SET_NETDEV_DEV function. It should register with the device hierarchy in two additional cases: standard (non-PnP) ISA and EISA.

- Currently they appear here:
/sys/devices/virtual/net/eth0 (standard ISA)
/sys/devices/virtual/net/eth1 (EISA)

Quoted 5 lines

- Rather, they should instead be here:
/sys/devices/isa/3c509.0/net/eth0 (standard ISA) /sys/devices/pci0000:00/0000:00:07.0/00:04/net/eth1 (EISA)

Tested on ISA and EISA boards.

That's good but you forgot to sign off, so the patch can't be applied.

BTW, it seems that after almost 7 years I seem to have found a tester for
my patch which lingered all this time in the Andrew Morton's tree! You've said you have EISA boards, how about testing my patch? Now if I could find Andrew Morton's tree... Andrew?
The patch is for 3c59x.c however -- I've just realized the difference.
Just in case, here's the link to 2008 version of it:

http://marc.info/?l=linux-netdev&m=121744737806819

WBR, Sergei

#2 lore raw
Quoted 22 lines · On Fri, 26 Apr 2013 01:03:53 +0400 Sergei Shtylyov…

Hello.

Quoted 10 lines · Matthew Whitehead

From: Matthew Whitehead <[email protected]>

The venerable 3c509 driver only sets its device parent in one case, the ISAPnP one. It does this with the SET_NETDEV_DEV function. It should register with the device hierarchy in two additional cases: standard (non-PnP) ISA and EISA.

- Currently they appear here:
/sys/devices/virtual/net/eth0 (standard ISA)
/sys/devices/virtual/net/eth1 (EISA)

Quoted 5 lines

- Rather, they should instead be here:
/sys/devices/isa/3c509.0/net/eth0 (standard ISA) /sys/devices/pci0000:00/0000:00:07.0/00:04/net/eth1 (EISA)

Tested on ISA and EISA boards.

That's good but you forgot to sign off, so the patch can't be applied.

Quoted 8 lines

BTW, it seems that after almost 7 years I seem to have found a tester for
my patch which lingered all this time in the Andrew Morton's tree! You've said you have EISA boards, how about testing my patch? Now if I could find Andrew Morton's tree... Andrew?
The patch is for 3c59x.c however -- I've just realized the difference.
Just in case, here's the link to 2008 version of it:

http://marc.info/?l=linux-netdev&m=121744737806819

It looks like I dropped that patch two years ago ;)

From: Sergei Shtylyov <[email protected]>

The driver wrongly claimed I/O ports at an address returned by pci_iomap() -- even if it was passed an MMIO address. Fix this by claiming/releasing all PCI resources in the PCI driver probe/remove handlers instead and get rid of the must_free_region flag weirdness (why would Cardbus claim anything for us?).

Also, the remove handler was trying to talk to the chip after having disabled its address decoders (at least on x86) -- fix this and get rid of useless VORTEX_PCI() calls.

While at it, fix some cases of the overly indented code...

[[email protected]: coding-style fixes]
Cc Steffen Klassert <[email protected]>

---

 drivers/net/3c59x.c |   66 ++++++++++++++++++++----------------------
 1 file changed, 32 insertions(+), 34 deletions(-)
drivers/net/3c59x.c renamed from drivers/net/3c59x.c~3x59x-fix-pci-resource-management +32 -34
diff -puN drivers/net/3c59x.c~3x59x-fix-pci-resource-management drivers/net/3c59x.c
@@ -632,7 +632,6 @@ struct vortex_private {
 		pm_state_valid:1,				/* pci_dev->saved_config_space has sane contents */
 		open:1,
 		medialock:1,
-		must_free_region:1,				/* Flag: if zero, Cardbus owns the I/O region */
 		large_frames:1;			/* accept large frames */
 	int drv_flags;
 	u16 status_enable;
@@ -1004,7 +1003,13 @@ static int __devinit vortex_init_one(str
 	/* wake up and enable device */
 	rc = pci_enable_device(pdev);
 	if (rc < 0)
-		goto out;
+		return rc;
+
+	rc = pci_request_regions(pdev, "3c59x");
+	if (rc < 0) {
+		pci_disable_device(pdev);
+		return rc;
+	}
 
 	unit = vortex_cards_found;
 
@@ -1024,13 +1029,13 @@ static int __devinit vortex_init_one(str
 	rc = vortex_probe1(&pdev->dev, ioaddr, pdev->irq,
 			   ent->driver_data, unit);
 	if (rc < 0) {
+		pci_release_regions(pdev);
 		pci_disable_device(pdev);
-		goto out;
+		return rc;
 	}
 
 	vortex_cards_found++;
 
-out:
 	return rc;
 }
 
@@ -1168,11 +1173,6 @@ static int __devinit vortex_probe1(struc
 
 	/* PCI-only startup logic */
 	if (pdev) {
-		/* EISA resources already marked, so only PCI needs to do this here */
-		/* Ignore return value, because Cardbus drivers already allocate for us */
-		if (request_region(dev->base_addr, vci->io_size, print_name) != NULL)
-			vp->must_free_region = 1;
-
 		/* enable bus-mastering if necessary */
 		if (vci->flags & PCI_USES_MASTER)
 			pci_set_master(pdev);
@@ -1205,12 +1205,13 @@ static int __devinit vortex_probe1(struc
 	vp->mii.reg_num_mask = 0x1f;
 
 	/* Makes sure rings are at least 16 byte aligned. */
-	vp->rx_ring = pci_alloc_consistent(pdev, sizeof(struct boom_rx_desc) * RX_RING_SIZE
-					   + sizeof(struct boom_tx_desc) * TX_RING_SIZE,
-					   &vp->rx_ring_dma);
+	vp->rx_ring = pci_alloc_consistent(pdev,
+				sizeof(struct boom_rx_desc) * RX_RING_SIZE +
+				sizeof(struct boom_tx_desc) * TX_RING_SIZE,
+				&vp->rx_ring_dma);
 	retval = -ENOMEM;
 	if (!vp->rx_ring)
-		goto free_region;
+		goto free_device;
 
 	vp->tx_ring = (struct boom_tx_desc *)(vp->rx_ring + RX_RING_SIZE);
 	vp->tx_ring_dma = vp->rx_ring_dma + sizeof(struct boom_rx_desc) * RX_RING_SIZE;
@@ -1471,13 +1472,11 @@ static int __devinit vortex_probe1(struc
 
 free_ring:
 	pci_free_consistent(pdev,
-						sizeof(struct boom_rx_desc) * RX_RING_SIZE
-							+ sizeof(struct boom_tx_desc) * TX_RING_SIZE,
-						vp->rx_ring,
-						vp->rx_ring_dma);
-free_region:
-	if (vp->must_free_region)
-		release_region(dev->base_addr, vci->io_size);
+			    sizeof(struct boom_rx_desc) * RX_RING_SIZE +
+			    sizeof(struct boom_tx_desc) * TX_RING_SIZE,
+			    vp->rx_ring,
+			    vp->rx_ring_dma);
+free_device:
 	free_netdev(dev);
 	pr_err(PFX "vortex_probe1 fails.  Returns %d\n", retval);
 out:
@@ -3169,29 +3168,28 @@ static void __devexit vortex_remove_one(
 	vp = netdev_priv(dev);
 
 	if (vp->cb_fn_base)
-		pci_iounmap(VORTEX_PCI(vp), vp->cb_fn_base);
+		pci_iounmap(pdev, vp->cb_fn_base);
 
 	unregister_netdev(dev);
 
-	if (VORTEX_PCI(vp)) {
-		pci_set_power_state(VORTEX_PCI(vp), PCI_D0);	/* Go active */
-		if (vp->pm_state_valid)
-			pci_restore_state(VORTEX_PCI(vp));
-		pci_disable_device(VORTEX_PCI(vp));
-	}
+	pci_set_power_state(pdev, PCI_D0);	/* Go active */
+	if (vp->pm_state_valid)
+		pci_restore_state(pdev);
+
 	/* Should really use issue_and_wait() here */
 	iowrite16(TotalReset | ((vp->drv_flags & EEPROM_RESET) ? 0x04 : 0x14),
 	     vp->ioaddr + EL3_CMD);
 
-	pci_iounmap(VORTEX_PCI(vp), vp->ioaddr);
+	pci_iounmap(pdev, vp->ioaddr);
 
 	pci_free_consistent(pdev,
-						sizeof(struct boom_rx_desc) * RX_RING_SIZE
-							+ sizeof(struct boom_tx_desc) * TX_RING_SIZE,
-						vp->rx_ring,
-						vp->rx_ring_dma);
-	if (vp->must_free_region)
-		release_region(dev->base_addr, vp->io_size);
+			    sizeof(struct boom_rx_desc) * RX_RING_SIZE +
+			    sizeof(struct boom_tx_desc) * TX_RING_SIZE,
+			    vp->rx_ring,
+			    vp->rx_ring_dma);
+
+	pci_release_regions(pdev);
+	pci_disable_device(pdev);
 	free_netdev(dev);
 }
 

_

#3 lore raw
In reply to Sergei Shtylyov

Hello.

Quoted 16 lines · Sergei Shtylyov
Quoted 15 lines

From: Matthew Whitehead <[email protected]>

The venerable 3c509 driver only sets its device parent in one case, the ISAPnP one.
It does this with the SET_NETDEV_DEV function. It should register with the device
hierarchy in two additional cases: standard (non-PnP) ISA and EISA.

- Currently they appear here:
/sys/devices/virtual/net/eth0 (standard ISA)
/sys/devices/virtual/net/eth1 (EISA)

- Rather, they should instead be here:
/sys/devices/isa/3c509.0/net/eth0 (standard ISA) /sys/devices/pci0000:00/0000:00:07.0/00:04/net/eth1 (EISA)

Quoted 8 lines
Quoted 1 line

Tested on ISA and EISA boards.

That's good but you forgot to sign off, so the patch can't be applied.

BTW, it seems that after almost 7 years I seem to have found a tester for
my patch which lingered all this time in the Andrew Morton's tree! You've said you have EISA boards, how about testing my patch? Now if I could find Andrew Morton's tree... Andrew?

Found it at last at ozlabs.org but it turned out that Andrew has dropped
the patch without notifying me. :-(

WBR, Sergei

#4 lore raw
Quoted 24 lines · On Fri, 26 Apr 2013 01:24:15 +0400 Sergei Shtylyov…

Hello.

Quoted 19 lines · Sergei Shtylyov
Quoted 12 lines

From: Matthew Whitehead <[email protected]>

The venerable 3c509 driver only sets its device parent in one case, the ISAPnP one.
It does this with the SET_NETDEV_DEV function. It should register with the device
hierarchy in two additional cases: standard (non-PnP) ISA and EISA.

- Currently they appear here:
/sys/devices/virtual/net/eth0 (standard ISA)
/sys/devices/virtual/net/eth1 (EISA)

Quoted 5 lines

- Rather, they should instead be here:
/sys/devices/isa/3c509.0/net/eth0 (standard ISA) /sys/devices/pci0000:00/0000:00:07.0/00:04/net/eth1 (EISA)

Tested on ISA and EISA boards.

That's good but you forgot to sign off, so the patch can't be applied.

Quoted 8 lines
Quoted 4 lines

BTW, it seems that after almost 7 years I seem to have found a tester for
my patch which lingered all this time in the Andrew Morton's tree! You've said you have EISA boards, how about testing my patch? Now if I could find Andrew Morton's tree... Andrew?

Found it at last at ozlabs.org but it turned out that Andrew has dropped
the patch without notifying me. :-(

nah, you were asleep ;)

Message-Id: <[email protected]>
Subject: [obsolete] 3x59x-fix-pci-resource-management.patch removed from -mm tree
To: [email protected], [email protected]
From: [email protected]
Date: Wed, 10 Jun 2009 12:55:21 -0700

#5 lore raw

Hello.

Quoted 9 lines · Andrew Morton
Quoted 7 lines
Quoted 4 lines

BTW, it seems that after almost 7 years I seem to have found a tester for
my patch which lingered all this time in the Andrew Morton's tree! You've said you have EISA boards, how about testing my patch? Now if I could find Andrew Morton's tree... Andrew?

Found it at last at ozlabs.org but it turned out that Andrew has dropped
the patch without notifying me. :-(

nah, you were asleep ;)

Nope, I was actively protesting, although I don't have my messages now,
have only found yours.

Quoted 5 lines

Message-Id: <[email protected]>
Subject: [obsolete] 3x59x-fix-pci-resource-management.patch removed from -mm tree
To: [email protected], [email protected]
From: [email protected]
Date: Wed, 10 Jun 2009 12:55:21 -0700

WBR, Sergei

#6 lore raw

Hello.

Quoted 7 lines · Sergei Shtylyov
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

BTW, it seems that after almost 7 years I seem to have found a tester
for
my patch which lingered all this time in the Andrew Morton's tree! You've said
you have EISA boards, how about testing my patch? Now if I could find Andrew Morton's tree... Andrew?

Quoted 14 lines
Quoted 4 lines
Quoted 2 lines

Found it at last at ozlabs.org but it turned out that Andrew has dropped
the patch without notifying me. :-(

nah, you were asleep ;)

Nope, I was actively protesting, although I don't have my messages now,
have only found yours.

Quoted 6 lines

Message-Id: <[email protected]>
Subject: [obsolete] 3x59x-fix-pci-resource-management.patch removed from -mm tree
To: [email protected], [email protected]
From: [email protected]
Date: Wed, 10 Jun 2009 12:55:21 -0700

In fact, you've readded the patch top you tree the next day. Then the
discussion continued in August 2009 until DaveM told you to repost this once it had at least CardBus testing...

WBR, Sergei