[PATCH][Take2] PCI legacy I/O port free driver - Making Intel e1000 driver legacy I/O port free

[Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com>]

Dear Auke

http://lkml.org/lkml/2007/5/16/275
> I'm ok with the bottom part of the patch, but I do not like the modification of
> the pci device ID table in this way. As Arjan van der Ven previously commented
> as well, this makes it hard for future device ID's to be bound to the driver.
>
> On top of that, there is no logical correlation between the mapping and
> chipsets, so a lot of information is lost in that table. It really does not show
> which _chipsets_ support this functionality.
>
> I think if we want to work with this, we need some way of mapping the device
> ID's back to chipsets, and enable the feature on that basis.

 This patches will meet your need in the way that it shows correlation
 between their device IDs and the chipsets. It does not use the existing
 macro INTEL_E1000_ETHERNET_DEVICE() to see whether the device uses I/O
 port or not. Instead of modifying the PCI device table, I've added another
 function called e1000_test_legacy_ioport() which tells you whether the PCI
 device uses legacy I/O port or not by checking its chipset.

 But one thing I want to say is that I am not sure about the chipsets that
 require legacy I/O port. I found the following code in e1000 driver code
 and it seems to be the only part that is using I/O port. So I thought
 the following chipsets are the only ones using legacy I/O port.
 I might be wrong, so any comments would be helpful.

    drivers/net/e1000/e1000_hw.c
     524 int32_t
     525 e1000_reset_hw(struct e1000_hw *hw)
     526 {
     ...
     618     switch (hw->mac_type) {
     619         case e1000_82544:
     620         case e1000_82540:
     621         case e1000_82545:
     622         case e1000_82546:
     623         case e1000_82541:
     624         case e1000_82541_rev_2:
     625             /* These controllers can't ack the 64-bit write when issuing the
     626              * reset, so use IO-mapping as a workaround to issue the reset */
     627             E1000_WRITE_REG_IO(hw, CTRL, (ctrl | E1000_CTRL_RST));

> I also would like this option to be non-default, IOW use legacy IO by default,
> and allow the user to specify a module load option to disable use of this feature:

 I've also added the module parameter so that the user can select whether
 he or she wants to enable the legacy I/O port free. The legacy I/O port free
 option is non-default.

 Rest of the part has not been changed since my previous patch. Any comments
 would be helpful.

Tomohiro Kusumi

Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com>

---
diff -Nurp linux-2.6.22.org/drivers/net/e1000/e1000.h linux-2.6.22/drivers/net/e1000/e1000.h
--- linux-2.6.22.org/drivers/net/e1000/e1000.h	2007-07-09 08:32:17.000000000 +0900
+++ linux-2.6.22/drivers/net/e1000/e1000.h	2007-08-10 09:56:03.000000000 +0900
@@ -342,6 +342,9 @@ struct e1000_adapter {
 	boolean_t quad_port_a;
 	unsigned long flags;
 	uint32_t eeprom_wol;
+
+	int use_ioport;
+	int bars;
 };

 enum e1000_state_t {
diff -Nurp linux-2.6.22.org/drivers/net/e1000/e1000_main.c linux-2.6.22/drivers/net/e1000/e1000_main.c
--- linux-2.6.22.org/drivers/net/e1000/e1000_main.c	2007-07-09 08:32:17.000000000 +0900
+++ linux-2.6.22/drivers/net/e1000/e1000_main.c	2007-08-10 13:09:25.000000000 +0900
@@ -222,6 +222,11 @@ static pci_ers_result_t e1000_io_error_d
 static pci_ers_result_t e1000_io_slot_reset(struct pci_dev *pdev);
 static void e1000_io_resume(struct pci_dev *pdev);

+static unsigned int enable_legacy_ioport_free = 0;
+module_param(enable_legacy_ioport_free, uint, 0644);
+MODULE_PARM_DESC(enable_legacy_ioport_free, "Enable legacy I/O port free (default:0)");
+static int e1000_test_legacy_ioport(struct pci_dev *pdev);
+
 static struct pci_error_handlers e1000_err_handler = {
 	.error_detected = e1000_io_error_detected,
 	.slot_reset = e1000_io_slot_reset,
@@ -868,8 +873,22 @@ e1000_probe(struct pci_dev *pdev,
 	int i, err, pci_using_dac;
 	uint16_t eeprom_data = 0;
 	uint16_t eeprom_apme_mask = E1000_EEPROM_APME;
-	if ((err = pci_enable_device(pdev)))
+	int bars = 0;
+	int use_ioport = 0;
+
+	if (enable_legacy_ioport_free) {
+		if ((use_ioport = e1000_test_legacy_ioport(pdev)) < 0) {
+			E1000_ERR("e1000_test_legacy_ioport failed, aborting\n");
+			return -1;
+		}
+		if (use_ioport)
+			bars = pci_select_bars(pdev, IORESOURCE_MEM | IORESOURCE_IO);
+		else
+			bars = pci_select_bars(pdev, IORESOURCE_MEM);
+	}
+	else if ((err = pci_enable_device(pdev))) {
 		return err;
+	}

 	if (!(err = pci_set_dma_mask(pdev, DMA_64BIT_MASK)) &&
 	    !(err = pci_set_consistent_dma_mask(pdev, DMA_64BIT_MASK))) {
@@ -883,7 +902,8 @@ e1000_probe(struct pci_dev *pdev,
 		pci_using_dac = 0;
 	}

-	if ((err = pci_request_regions(pdev, e1000_driver_name)))
+	if ((enable_legacy_ioport_free && (err = pci_request_selected_regions(pdev, bars, e1000_driver_name))) ||
+	    (err = pci_request_regions(pdev, e1000_driver_name)))
 		goto err_pci_reg;

 	pci_set_master(pdev);
@@ -902,6 +922,10 @@ e1000_probe(struct pci_dev *pdev,
 	adapter->pdev = pdev;
 	adapter->hw.back = adapter;
 	adapter->msg_enable = (1 << debug) - 1;
+	if (enable_legacy_ioport_free) {
+		adapter->use_ioport = use_ioport;
+		adapter->bars = bars;
+	}

 	mmio_start = pci_resource_start(pdev, BAR_0);
 	mmio_len = pci_resource_len(pdev, BAR_0);
@@ -911,12 +935,14 @@ e1000_probe(struct pci_dev *pdev,
 	if (!adapter->hw.hw_addr)
 		goto err_ioremap;

-	for (i = BAR_1; i <= BAR_5; i++) {
-		if (pci_resource_len(pdev, i) == 0)
-			continue;
-		if (pci_resource_flags(pdev, i) & IORESOURCE_IO) {
-			adapter->hw.io_base = pci_resource_start(pdev, i);
-			break;
+	if (!enable_legacy_ioport_free || adapter->use_ioport) {
+		for (i = BAR_1; i <= BAR_5; i++) {
+			if (pci_resource_len(pdev, i) == 0)
+				continue;
+			if (pci_resource_flags(pdev, i) & IORESOURCE_IO) {
+				adapter->hw.io_base = pci_resource_start(pdev, i);
+				break;
+			}
 		}
 	}

@@ -1182,7 +1208,10 @@ err_sw_init:
 err_ioremap:
 	free_netdev(netdev);
 err_alloc_etherdev:
-	pci_release_regions(pdev);
+	if (enable_legacy_ioport_free)
+		pci_release_selected_regions(pdev, bars);
+	else
+		pci_release_regions(pdev);
 err_pci_reg:
 err_dma:
 	pci_disable_device(pdev);
@@ -1234,7 +1263,10 @@ e1000_remove(struct pci_dev *pdev)
 	iounmap(adapter->hw.hw_addr);
 	if (adapter->hw.flash_address)
 		iounmap(adapter->hw.flash_address);
-	pci_release_regions(pdev);
+	if (enable_legacy_ioport_free)
+		pci_release_selected_regions(pdev, adapter->bars);
+	else
+		pci_release_regions(pdev);

 	free_netdev(netdev);

@@ -5175,7 +5207,8 @@ e1000_resume(struct pci_dev *pdev)

 	pci_set_power_state(pdev, PCI_D0);
 	pci_restore_state(pdev);
-	if ((err = pci_enable_device(pdev))) {
+	if ((enable_legacy_ioport_free && (err = pci_enable_device_bars(pdev, adapter->bars))) ||
+	    (err = pci_enable_device(pdev))) {
 		printk(KERN_ERR "e1000: Cannot enable PCI device from suspend\n");
 		return err;
 	}
@@ -5320,4 +5353,43 @@ static void e1000_io_resume(struct pci_d

 }

+/*
+ * e1000_test_legacy_ioport - see if the device uses legacy I/O port
+ * @pdev: Pointer to PCI device
+ *
+ * This function tests if the PCI device uses I/O port.
+ * If yes the function returns 1, if no the function returns 0.
+ * The function returns -1 if there is an error.
+ */
+
+static int e1000_test_legacy_ioport(struct pci_dev *pdev)
+{
+	int ret;
+	struct e1000_hw hw;
+	memset(&hw, 0, sizeof(hw));
+	
+	hw.mac_type = e1000_undefined;
+	hw.device_id = pdev->device;
+	pci_read_config_byte(pdev, PCI_REVISION_ID, &hw.revision_id);
+
+	if (e1000_set_mac_type(&hw) < 0)
+		return -1;
+
+	switch (hw.mac_type) {
+	case e1000_82540:
+	case e1000_82541:
+	case e1000_82541_rev_2:
+	case e1000_82544:
+	case e1000_82545:
+	case e1000_82546:
+		ret = 1;
+		break;
+	default:
+		ret = 0;
+		break;
+	}
+	
+	return ret;
+}
+
 /* e1000_main.c */


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/