RE: [PATCH 0/3] ARM: OMAP: SmartReflex driver

[Kalle Jokiniemi <ext-kalle.jokiniemi@nokia.com>]

Hi Pratheesh,

On ma, 2008-06-02 at 20:28 +0530, ext TK, Pratheesh Gangadhar wrote:
> Koen,
> 
> You need to use NTargets from E-fuse for the correct operation. If E-fuse reads zero, then you are using a device without SmartReflex support or not guaranteed at least.
> 
> #define CONTROL_FUSE_OPP1_VDD1           0x48002380
> #define CONTROL_FUSE_OPP2_VDD1           0x48002384
> #define CONTROL_FUSE_OPP3_VDD1           0x48002388
> #define CONTROL_FUSE_OPP4_VDD1           0x4800238C
> #define CONTROL_FUSE_OPP5_VDD1           0x48002390
> 
> #define CONTROL_FUSE_OPP1_VDD2           0x48002394
> #define CONTROL_FUSE_OPP2_VDD2           0x48002398
> #define CONTROL_FUSE_OPP3_VDD2           0x4800239C
> 
> #define CONTROL_FUSE_SR                  0x480023A0
> 
> Those values present in the driver are meant for testing purposes only. Ideally in sr_enable, if current_nvalue reads zero we should return without enabling SmartReflex module.
> 
> 
>         current_nvalue = sr_read_reg(sr, NVALUERECIPROCAL);
> 
>         if (nvalue_reciprocal == 0) {
>                 DPRINTK("OPP doesn't support SmartReflex\n");
>                 return;
>         }
> 
>         if (current_nvalue == nvalue_reciprocal) {
>                 DPRINTK("System is already at the desired voltage level\n");
>                 return;
>         }

This last if-statement is something that puzzled me while I was doing
this patch-set. Shouldn't the SmartReflex be enabled even though we are
at the desired voltage level at a given time. I mean, the operating
conditions (temperature for one) could change over time and smartreflex
will need to do it's magic in order to keep the device at a proper
voltage level, right? 


> 
> This driver needs modification to read NTargets from E-fuse rather than using hard coded values.

Tony, I can update this patch set, or send an additional patch that adds
the efuse support. Which will you prefer?


Regards,
Kalle


> -----Original Message-----
> From: linux-omap-owner@vger.kernel.org [mailto:linux-omap-owner@vger.kernel.org] On Behalf Of Koen Kooi
> Sent: Monday, June 02, 2008 6:02 PM
> To: Kalle Jokiniemi
> Cc: linux-omap
> Subject: Re: [PATCH 0/3] ARM: OMAP: SmartReflex driver
> 
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> 
> Op 2 jun 2008, om 13:29 heeft Kalle Jokiniemi het volgende geschreven:
> 
> > Here is an updated version of the SmartReflex driver.
> >
> > Patches 1 and 2 are same as before, but patch 3 has some new things:
> >
> > - Changed the register accessing to use the prm_xxx_mod_reg{_bits}()
> >  functions.
> >
> > - Created a Kconfig entry for SmartReflex in "System type->TI OMAP
> >  Implementations" menu.
> >
> > Tested on 3430SDP.
> 
> Compiles for beagleboard, but the board hangs after:
> 
> echo -n 1 > /sys/power/sr_vdd1_autocomp
> echo -n 1 > /sys/power/sr_vdd2_autocomp
> 
> But that's likely to be a beagleboard specific problem.
> 
> regards,
> 
> Koen
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.5 (Darwin)
> 
> iD8DBQFIQ+g/MkyGM64RGpERAqwgAJ97AE4z0bIBZJs7nZcO1Npa8YkIcgCfY6KM
> 3Hhr4AIdmzNiozdqs1uQUgg=
> =rcQP
> -----END PGP SIGNATURE-----
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html