| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Fri, 07 Jun 2013 10:47:37 +0200
From: Arnd Bergmann <arnd@...db.de>
To: Alexey Brodkin <Alexey.Brodkin@...opsys.com>
Cc: netdev@...r.kernel.org, Vineet Gupta <Vineet.Gupta1@...opsys.com>,
Mischa Jonker <Mischa.Jonker@...opsys.com>,
Grant Likely <grant.likely@...aro.org>,
Rob Herring <rob.herring@...xeda.com>,
Paul Gortmaker <paul.gortmaker@...driver.com>,
"David S. Miller" <davem@...emloft.net>,
"Steven J. Hill" <sjhill@...s.com>, linux-kernel@...r.kernel.org,
devicetree-discuss@...ts.ozlabs.org
Subject: Re: [PATCH] ethernet/arc/arc_emac - Add new driver
On Tuesday 04 June 2013 16:21:50 Alexey Brodkin wrote:
> drivers/net/ethernet/Kconfig | 1 +
> drivers/net/ethernet/Makefile | 1 +
> drivers/net/ethernet/arc/Kconfig | 29 +
> drivers/net/ethernet/arc/Makefile | 6 +
> drivers/net/ethernet/arc/arc_emac_main.c | 905 ++++++++++++++++++++++++++++++
> drivers/net/ethernet/arc/arc_emac_main.h | 82 +++
> drivers/net/ethernet/arc/arc_emac_mdio.c | 181 ++++++
> drivers/net/ethernet/arc/arc_emac_mdio.h | 22 +
> drivers/net/ethernet/arc/arc_emac_regs.h | 73 +++
I wonder if it would be better to name the directory "synopsys" or
"designware" rather than "arc" now. Is there a chance that the same
controller is used on non-arc CPUs?
> +static int arc_emac_probe(struct platform_device *pdev)
> +{
> + struct net_device *net_dev;
> + struct arc_emac_priv *priv;
> + int err;
> + unsigned int clock_frequency;
> + unsigned int id;
> + struct resource res_regs;
> +#ifdef CONFIG_OF_IRQ
> + struct resource res_irq;
> +#endif
> + const char *mac_addr = NULL;
Please remove the #ifdef here. The driver does not work without this
anyway, so better make it 'depend on OF_IRQ' in Kconfig.
> + /* Get phy from device tree */
> + priv->phy_node = of_parse_phandle(pdev->dev.of_node, "phy", 0);
> + if (!priv->phy_node) {
> + dev_err(&pdev->dev,
> + "failed to retrieve phy description from device tree\n");
> + err = -ENODEV;
> + goto out;
> + }
You should add a binding document in Documentation/devicetree/bindings that
describes what properties are required.
> + /* Get EMAC registers base address from device tree */
> + err = of_address_to_resource(pdev->dev.of_node, 0, &res_regs);
> + if (err) {
> + dev_err(&pdev->dev,
> + "failed to retrieve base register from device tree\n");
> + err = -ENODEV;
> + goto out;
> + }
> +
> + if (!devm_request_mem_region(&pdev->dev, res_regs.start,
> + resource_size(&res_regs), pdev->name)) {
> + dev_err(&pdev->dev,
> + "failed to request memory region for base registers\n");
> + err = -ENXIO;
> + goto out;
> + }
> +
> + priv->reg_base_addr = (void *) devm_ioremap_nocache(&pdev->dev,
> + res_regs.start, resource_size(&res_regs));
> + if (!priv->reg_base_addr) {
> + dev_err(&pdev->dev, "failed to ioremap MAC registers\n");
> + err = -ENXIO;
> + goto out;
> + }
This block can be simplified to a single devm_ioremap_resource() now.
The cast to 'void *' is wrong: please make sure that you always use '__iomem'
pointers for MMIO mappings. You can use 'sparse' with 'make C=1' to check this
at build time.
> + /* Get MAC address from device tree */
> +#ifdef CONFIG_OF_NET
> + mac_addr = of_get_mac_address(pdev->dev.of_node);
> +#endif
Same as above, remove the #ifdef.
> diff --git a/drivers/net/ethernet/arc/arc_emac_main.h b/drivers/net/ethernet/arc/arc_emac_main.h
> new file mode 100644
> index 0000000..6f03d26
> --- /dev/null
> +++ b/drivers/net/ethernet/arc/arc_emac_main.h
This header seems to be included only in the main .c file, just
move the contents there and remove this file.
> +/**
> + * arc_mdio_probe - MDIO probe function.
> + * @dev_node: Pointer to device node.
> + * @priv: Pointer to ARC MDIO private data structure.
> + *
> + * returns: 0 on success, -ENOMEM when mdiobus_alloc
> + * (to allocate memory for MII bus structure) fails.
> + *
> + * Sets up and registers the MDIO interface.
> + */
> +int arc_mdio_probe(struct device_node *dev_node, struct arc_mdio_priv *priv)
> +{
> + struct device_node *mdio_np;
> + struct mii_bus *bus;
> + int error;
> +
> + bus = mdiobus_alloc();
> + if (!bus) {
> + error = -ENOMEM;
> + goto cleanup;
> + }
> +
> + priv->bus = bus;
> + bus->priv = priv;
> + bus->name = "Synopsys MII Bus",
> + bus->read = &arc_mdio_read;
> + bus->write = &arc_mdio_write;
> +
> + snprintf(bus->id, MII_BUS_ID_SIZE, "%.8x",
> + (unsigned int)priv->reg_base_addr);
> +
> + bus->parent = priv->dev;
> +
> + mdio_np = of_find_node_by_name(NULL, "mdio");
> + if (!mdio_np) {
> + dev_err(priv->dev, "cannot find <mdio> in device tree\n");
> + error = -ENODEV;
> + goto cleanup;
> + }
of_find_node_by_name() is probably not what you want here, the name should
not be used as a primary key. Maybe it's better to use a standalone driver
for the phy and put it into drivers/net/phy/. I don't know what the official
policy is here though, since the phy is only used in this one driver.
Arnd
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists