On Wed, Jan 4, 2017 at 4:21 AM, Phil Reid <preid@xxxxxxxxxxxxxxxxx> wrote: > G'day Andy, > > On 4/01/2017 07:05, Andy Shevchenko wrote: >> >> On Wed, Jan 4, 2017 at 1:02 AM, Andy Shevchenko >> <andy.shevchenko@xxxxxxxxx> wrote: >>> >>> On Tue, Jan 3, 2017 at 6:54 AM, Phil Reid <preid@xxxxxxxxxxxxxxxxx> >>> wrote: >>>> >>>> Some system have multiple dw devices. Currently the driver uses a >>>> fixed name for the debugfs dir. Append bus_num to the debugfs dir >>>> name to make it unique. >>>> >>> >>> Reviewed-by: Andy Shevchenko <andy.shevchenko@xxxxxxxxx> >> >> >> Ah, hold on. >> dws->name is exactly what you are trying to re-invent here. Just use >> it directly. >> > I did try that initially. > However when I used that they where all named dw_spi65535. > > This seems to be because bus_num is set to -1 ion the call to > dw_spi_add_host. > Which results in dynamic bus_num assignment in the call to > devm_spi_register_master. > > Moving 'snprintf(dws->name, sizeof(dws->name), "dw_spi%d", dws->bus_num);' > to after > devm_spi_register_master results in a warning being emitting when > request_irq is called > about an emit string at : > WARNING: CPU: 0 PID: 1 at fs/proc/generic.c:369 __proc_create+0x1e8/0x1f0 > > cat /proc/interrupts currently also shows dw_spi65535 > > Rewrite dws->name after devm_spi_register_master is called fixes > /proc/interrupts > And it could then be used in debugfs. > But does not fix the filename in /proc/irq/<num>/dw_spi65535 > > I couldn't see a safe workaround with dws->name > Unless we can move the request_irq to after devm_spi_register_master > > Thoughts? I briefly checked couple of SPI drivers, they are doing request_irq(..., dev_name(...), ...); So, I propose to use that pattern for now, and move snprintf() call to debugfs related code. -- With Best Regards, Andy Shevchenko -- To unsubscribe from this list: send the line "unsubscribe linux-spi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html