Messages in this thread | | | Date | Wed, 20 Feb 2013 09:53:50 +0100 | Subject | Re: [PATCH 06/35] mfd: ab8500-core: Sysfs chip id modification | From | Marcus Cooper <> |
| |
On 20 February 2013 09:13, Lee Jones <lee.jones@linaro.org> wrote: > On Tue, 19 Feb 2013, Arnd Bergmann wrote: > >> On Friday 15 February 2013, Lee Jones wrote: >> > struct ab8500 *ab8500; >> > + int chip_id = -EINVAL; >> > >> > ab8500 = dev_get_drvdata(dev); >> > - return sprintf(buf, "%#x\n", ab8500 ? ab8500->chip_id : -EINVAL); >> > + if(ab8500) { >> > + chip_id = ab8500->chip_id; >> > + if((is_ab8505(ab8500) || is_ab9540(ab8500)) && ab8500->version != 0xFF) >> > + chip_id = (ab8500->version << 8) | chip_id; >> > + } >> > + return sprintf(buf, "%#x\n", chip_id); >> > } >> >> What's the use of printing "ffffffea" for unknown versions here? > > You mean instead of -EINVAL? No idea, Marcus? > Looks like I'm guilty of just making the minimal changes. Arnd is right though, getting ffffffea(-EINVAL) back is pretty useless. I'll have to check user land to see what is using this. Maybe not printing and returning 0 should be the correct behaviour when an unknown version is found.
| |