| Message ID | 20240902181530.6852-3-rosenp@gmail.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | net: ibm: emac: some cleanups and devm | expand |
On Mon, Sep 02, 2024 at 11:15:11AM -0700, Rosen Penev wrote: > It's the last to go in remove. Safe to let devm handle it. > > Also move request_irq to probe for clarity. It's removed in _remove not > close. > > Use dev_err instead of printk. Handles names automatically. > > + /* Setup error IRQ handler */ > + err = devm_request_irq(&ofdev->dev, dev->emac_irq, emac_irq, 0, "EMAC", dev); > + if (err) { > + dev_err(&ofdev->dev, "failed to request IRQ %d", dev->emac_irq); > + goto err_gone; > + } Is this an internal interrupt, or a GPIO? It could be it is done in open because there is a danger the GPIO controller has not probed yet. So here you might get an EPROBE_DEFFER, where as the much older kernel this was written for might not of done, if just gave an error had gave up. So dev_err_probe() might be better. Andrew
On Mon, Sep 2, 2024 at 1:06 PM Andrew Lunn <andrew@lunn.ch> wrote: > > On Mon, Sep 02, 2024 at 11:15:11AM -0700, Rosen Penev wrote: > > It's the last to go in remove. Safe to let devm handle it. > > > > Also move request_irq to probe for clarity. It's removed in _remove not > > close. > > > > Use dev_err instead of printk. Handles names automatically. > > > > + /* Setup error IRQ handler */ > > + err = devm_request_irq(&ofdev->dev, dev->emac_irq, emac_irq, 0, "EMAC", dev); > > + if (err) { > > + dev_err(&ofdev->dev, "failed to request IRQ %d", dev->emac_irq); > > + goto err_gone; > > + } > > Is this an internal interrupt, or a GPIO? It could be it is done in > open because there is a danger the GPIO controller has not probed > yet. So here you might get an EPROBE_DEFFER, where as the much older > kernel this was written for might not of done, if just gave an error > had gave up. So dev_err_probe() might be better. Good call on that. In my experience, I get these EPROBE_DEFER errors on OpenWrt's ath79 target (QCA MIPS) but not on PowerPC when trying to use GPIOs. Nevertheless it seems to be good practice to use dev_err_probe anyway. Will fix in v2. > > Andrew
> > Is this an internal interrupt, or a GPIO? It could be it is done in > > open because there is a danger the GPIO controller has not probed > > yet. So here you might get an EPROBE_DEFFER, where as the much older > > kernel this was written for might not of done, if just gave an error > > had gave up. So dev_err_probe() might be better. > Good call on that. In my experience, I get these EPROBE_DEFER errors > on OpenWrt's ath79 target (QCA MIPS) but not on PowerPC when trying to > use GPIOs. Nevertheless it seems to be good practice to use > dev_err_probe anyway. Will fix in v2. You might want to look at https://elixir.bootlin.com/linux/v6.10.7/source/drivers/net/ethernet/ibm/emac/core.c#L2418 and then replace it by correctly handling EPROBE_DEFER. As you said, an old driver, needing some cleanup. Andrew
diff --git a/drivers/net/ethernet/ibm/emac/core.c b/drivers/net/ethernet/ibm/emac/core.c index 348702f462bd..98d1b711969b 100644 --- a/drivers/net/ethernet/ibm/emac/core.c +++ b/drivers/net/ethernet/ibm/emac/core.c @@ -1228,18 +1228,10 @@ static void emac_print_link_status(struct emac_instance *dev) static int emac_open(struct net_device *ndev) { struct emac_instance *dev = netdev_priv(ndev); - int err, i; + int i; DBG(dev, "open" NL); - /* Setup error IRQ handler */ - err = request_irq(dev->emac_irq, emac_irq, 0, "EMAC", dev); - if (err) { - printk(KERN_ERR "%s: failed to request IRQ %d\n", - ndev->name, dev->emac_irq); - return err; - } - /* Allocate RX ring */ for (i = 0; i < NUM_RX_BUFF; ++i) if (emac_alloc_rx_skb(dev, i)) { @@ -1293,8 +1285,6 @@ static int emac_open(struct net_device *ndev) return 0; oom: emac_clean_rx_ring(dev); - free_irq(dev->emac_irq, dev); - return -ENOMEM; } @@ -1408,8 +1398,6 @@ static int emac_close(struct net_device *ndev) emac_clean_tx_ring(dev); emac_clean_rx_ring(dev); - free_irq(dev->emac_irq, dev); - netif_carrier_off(ndev); return 0; @@ -3082,6 +3070,14 @@ static int emac_probe(struct platform_device *ofdev) err = -ENODEV; goto err_gone; } + + /* Setup error IRQ handler */ + err = devm_request_irq(&ofdev->dev, dev->emac_irq, emac_irq, 0, "EMAC", dev); + if (err) { + dev_err(&ofdev->dev, "failed to request IRQ %d", dev->emac_irq); + goto err_gone; + } + ndev->irq = dev->emac_irq; /* Map EMAC regs */ @@ -3237,8 +3233,6 @@ static int emac_probe(struct platform_device *ofdev) err_irq_unmap: if (dev->wol_irq) irq_dispose_mapping(dev->wol_irq); - if (dev->emac_irq) - irq_dispose_mapping(dev->emac_irq); err_gone: /* if we were on the bootlist, remove us as we won't show up and * wake up all waiters to notify them in case they were waiting @@ -3284,9 +3278,6 @@ static void emac_remove(struct platform_device *ofdev) if (dev->wol_irq) irq_dispose_mapping(dev->wol_irq); - if (dev->emac_irq) - irq_dispose_mapping(dev->emac_irq); - } /* XXX Features in here should be replaced by properties... */
It's the last to go in remove. Safe to let devm handle it. Also move request_irq to probe for clarity. It's removed in _remove not close. Use dev_err instead of printk. Handles names automatically. Signed-off-by: Rosen Penev <rosenp@gmail.com> --- drivers/net/ethernet/ibm/emac/core.c | 27 +++++++++------------------ 1 file changed, 9 insertions(+), 18 deletions(-)