>>-----Original Message----- >>From: Kevin Hilman [mailto:khilman@xxxxxxxxxxxxxxxxxxx] >>Sent: Wednesday, August 25, 2010 3:52 AM >>To: Gopinath, Thara >>Cc: linux-omap@xxxxxxxxxxxxxxx; paul@xxxxxxxxx; Cousson, Benoit; Sripathy, Vishwanath; Sawant, Anand; >>Derrick, David >>Subject: Re: [PATCHv2 1/8] OMAP: PM: Allowing an early init of pm debugfs driver. >> >>On Tue, 2010-08-24 at 15:16 -0700, Kevin Hilman wrote: >>> Thara Gopinath <thara@xxxxxx> writes: >>> >>> > This patch changes the pm_db_init from arch initcall to a postcore >>> > initcall. With arch initcall, it is impossible for pm driver that >>> > gets initialized prior to this driver to use one of the >>> > pm debug fs entries during its init. Making it a postcore initcall >>> > ensures that this drver gets initialized early on before any pm >>> > drivers. >>> >>> Instead of tinkering with initcall ordering, how about calling the pm >>> debug init from pm.c >> >>ignore this comment, pm.c is a device_initcall, so wont solve your >>problem. >> >>But I'd still like to know what PM drivers are being initialized so >>early they need this to be a postcore_initcall. It is needed from voltage and smartreflex layers for adding the debugfs entries. But considering these inits have been moved to device_initcall or late_initcall may be This change is no longer relevant. Thanks for catching this. Regards Thara >> >>Kevin >> ��.n��������+%������w��{.n�����{�������ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
