On Thu, May 23, 2013 at 08:00:08PM +0800, Libo Chen wrote: > when i2c malloc faild, we should not try to call release_mem_region. > aovid this, convert them to devm_* API. > > Signed-off-by: Libo Chen <libo.chen@xxxxxxxxxx> First, you did it right. This patch should not be combined with 1/3. It is not a trivial cleanup, so I prefer to keep them seperated. BTW did you test these patches on real hardware? > @@ -1198,19 +1187,6 @@ static int i2c_pxa_probe(struct platform_device *dev) > #endif > return 0; > > -eadapt: > - if (!i2c->use_pio) > - free_irq(irq, i2c); > -ereqirq: > - clk_disable(i2c->clk); > - iounmap(i2c->reg_base); > -eremap: > - clk_put(i2c->clk); > -eclk: > - kfree(i2c); > -emalloc: > - release_mem_region(res->start, resource_size(res)); > - return ret; > } > > static int i2c_pxa_remove(struct platform_device *dev) > @@ -1218,15 +1194,6 @@ static int i2c_pxa_remove(struct platform_device *dev) > struct pxa_i2c *i2c = platform_get_drvdata(dev); > > i2c_del_adapter(&i2c->adap); > - if (!i2c->use_pio) > - free_irq(i2c->irq, i2c); > - > - clk_disable(i2c->clk); > - clk_put(i2c->clk); > - > - iounmap(i2c->reg_base); > - release_mem_region(i2c->iobase, i2c->iosize); > - kfree(i2c); You remove too much. Who disables the clock now?
Attachment:
signature.asc
Description: Digital signature
