| 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
| ||
|
Message-ID: <4881796E12491D4BB15146FE0209CE643F5CA32B@DE02WEMBXB.internal.synopsys.com>
Date: Sat, 8 Jun 2013 11:19:55 +0000
From: Alexey Brodkin <Alexey.Brodkin@...opsys.com>
To: Joe Perches <joe@...ches.com>
CC: Alexey Brodkin <Alexey.Brodkin@...opsys.com>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
Vineet Gupta <Vineet.Gupta1@...opsys.com>,
"Mischa Jonker" <Mischa.Jonker@...opsys.com>,
Arnd Bergmann <arnd@...db.de>,
"Grant Likely" <grant.likely@...aro.org>,
Rob Herring <rob.herring@...xeda.com>,
Paul Gortmaker <paul.gortmaker@...driver.com>,
"David S. Miller" <davem@...emloft.net>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"devicetree-discuss@...ts.ozlabs.org"
<devicetree-discuss@...ts.ozlabs.org>,
"romieu@...zoreil.com" <romieu@...zoreil.com>
Subject: Re: [PATCH v2] ethernet/arc/arc_emac - Add new driver
On 06/07/2013 10:13 PM, Joe Perches wrote:
> On Fri, 2013-06-07 at 19:07 +0400, Alexey Brodkin wrote:
>> Driver for non-standard on-chip ethernet device ARC EMAC 10/100,
>> instantiated in some legacy ARC (Synopsys) FPGA Boards such as
>> ARCAngel4/ML50x.
>
> trivial comments only:
>
>> diff --git a/drivers/net/ethernet/arc/arc_emac_main.c
b/drivers/net/ethernet/arc/arc_emac_main.c
>> + * Alexey Brodkin: June 2013
>> + * -Upsteaming
> []
>> + * Vineet Gupta: June 2011
>> + * -Issues when working with 64b cache line size
> []
>> + * Vineet Gupta: May 2010
>> + * -Reduced foot-print of the main ISR (handling for error cases
moved out
> []
>> + * Vineet Gupta: Nov 2009
>> + * -Unified NAPI and Non-NAPI Code.
> []
>> + * Vineet Gupta: Nov 2009
> []
>> + * Amit Bhor, Sameer Dhavale: 2004
>
> Does the internal changelog add anything useful?
Well, I think at this point it's more of honoring people who contributed
into this driver through past years.
I saw this kind of changelogs in other drivers like "arc_uart" that was
recently upstreamed:
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/drivers/tty/serial/arc_uart.c
So if I still want to name contributor should I just list their names?
>> +static int arc_emac_poll(struct napi_struct *napi, int budget)
>> +{
>> + struct net_device *net_dev = napi->dev;
>> + struct arc_emac_priv *priv = netdev_priv(net_dev);
>> + struct sk_buff *skb, *skbnew;
>> + unsigned int i, loop, len, info, work_done = 0;
>> +
>> + /* Loop thru the BD chain, but not always from 0.
>> + * Start from right after where we last saw a packet.
>> + */
>> + i = priv->last_rx_bd;
>> +
>> + for (loop = 0; loop < RX_BD_NUM; loop++) {
>> + i = (i + 1) & (RX_BD_NUM - 1);
>> +
>> + info = priv->rxbd[i].info;
>> +
>> + /* BD contains a packet for CPU to grab */
>> + if (likely((info & OWN_MASK) == FOR_CPU)) {
>
> You could reduce indentation a level for all
> the lines that follow by using
> if (unlikely(!(...)
> continue;
To be honest I thought about doing it, but personally I don't like
inverted logic - I'd say it makes code logic not that straight-forward
for understanding. Still it might be only my feeling and as I've got at
last to advises to act so - I'll definitely follow your proposal.
>> + /* Get a new SKB from stack */
>> + skbnew = netdev_alloc_skb(net_dev,
>> + net_dev->mtu +
>> + EMAC_BUFFER_PAD);
>> +
>> + if (!skbnew) {
>> + netdev_err(net_dev, "Out of memory, "
>> + "dropping packet\n");
>
> OOM messages aren't particularly useful,
> and coalesce format please...
I noticed that some drivers don't print error messages for this kind of
errors. My intention was to make possible problem troubleshooting a bit
easier. But if it is not supposed to print "standard" errors in console
but simply return proper error code from the function - I'm fine with it
- makes code smaller)
>> +static irqreturn_t arc_emac_intr(int irq, void *dev_instance)
>> +{
>> + struct net_device *net_dev = (struct net_device *)dev_instance;
>
> Don't need to cast void *
> netdev is a more common variable name
Good point - thanks.
>> + if (status & TXINT_MASK) {
>> + unsigned int i, info;
>> + struct sk_buff *skb;
> []
>> + if (status & TXCH_MASK) {
>> + priv->stats.tx_errors++;
>> + priv->stats.tx_aborted_errors++;
>> + netdev_err(priv->net_dev,
>> + "Tx chaining err! txbd_dirty = %u\n",
>> + priv->txbd_dirty);
>
> You already have a local net_dev.
> why not use that?
Good point - overengineered stuff)
>> +int arc_emac_open(struct net_device *net_dev)
>> +{
>> + struct arc_emac_priv *priv = netdev_priv(net_dev);
>> + struct arc_emac_bd_t *bd;
>> + struct sk_buff *skb;
>> + int i;
>> +
>> + if (!priv->phy_node) {
>> + netdev_err(net_dev, "arc_emac_open: phy device is absent\n");
>
> If you really need the function name,
> it's better to use "%s: ", __func__
Reminder from previous "printk". Will clean-up.
>> +int arc_emac_tx(struct sk_buff *skb, struct net_device *net_dev)
>> +{
> []
>> +tx_next_chunk:
>> +
>> + info = priv->txbd[priv->txbd_curr].info;
>> + if (likely((info & OWN_MASK) == FOR_CPU)) {
>
> if (unlikely(!(etc...)
> return NETDEV_TX_BUSY;
>
> and save an indent level
Same comment as above - will invert logic.
>> +/**
>
>> +int arc_emac_set_address(struct net_device *net_dev, void *p)
> []
>> + EMAC_REG_SET(priv->reg_base_addr, R_ADDRH,
>> + *(unsigned int *)&net_dev->dev_addr[4] & 0x0000ffff);
>
> That doesn't seem endian friendly.
Right. Need to revisit this one.
>> +static int arc_emac_probe(struct platform_device *pdev)
>> +{
> []
>> + /* 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");
>
> Coalesce formats please
Could you please clarify how should I format lines in question?
I'm a bit lost here.
>> + err = of_address_to_resource(pdev->dev.of_node, 0, &res_regs);
>> + if (err) {
>> + dev_err(&pdev->dev, "failed to retrieve registers base "
>> + "from device tree\n");
> []
>> + /* Get IRQ from device tree */
>> + err = of_irq_to_resource(pdev->dev.of_node, 0, &res_irq);
>> + if (!err) {
>> + dev_err(&pdev->dev, "failed to retrieve <irq> value "
>> + "from device tree\n");
> []
-Alexey
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists