I sent this patch (with the change in both activate_usb_autosuspend, suggest_usb_autosuspend and the since removed instance in urbnum.c), to the mailing list on 2010-05-26 (Subject: Nits).
I suppose my karma is way to low to push it into the code since Auke said "for now I don't care about this patch." Well, it is good that it finally got in. By the way, nothing ever checks the len variable so that assignment is unnecessary. /MF On Fri, 2010-08-06 at 17:28 +0200, Luca Tettamanti wrote: > (resending, it seems that my previous message got lost in the moderation > queue...) > > Hello, > I noticed (as in "my mouse stopped working") that the heuristic that > should skip USB input devices when enabling auto suspend is not working. > After staring at the code long enough I realized that there's something > wrong with: > > len = readlink(filename, linkto, sizeof(link) - 1); > > sizeof(link) returns the size of link(2), not the linkto char array :P > > -- >8 -- > readlink is failing because sizeof(link) returns the size of 'link' > which is a function, not the intended char buffer. > Fix the typo. > > Signed-off-by: Luca Tettamanti <[email protected]> > --- > usb.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/usb.c b/usb.c > index f3a1bef..44d3252 100644 > --- a/usb.c > +++ b/usb.c > @@ -52,7 +52,7 @@ void activate_usb_autosuspend(void) > /* skip usb input devices */ > sprintf(filename, "/sys/bus/usb/devices/%s/driver", > dirent->d_name); > memset(linkto, 0, sizeof(linkto)); > - len = readlink(filename, linkto, sizeof(link) - 1); > + len = readlink(filename, linkto, sizeof(linkto) - 1); > if (strstr(linkto, "usbhid")) > continue; > > > Luca > > _______________________________________________ > Power mailing list > [email protected] > http://www.bughost.org/mailman/listinfo/power _______________________________________________ Power mailing list [email protected] http://www.bughost.org/mailman/listinfo/power
