Hi Valdis,
Thanks for your remark.
Concerning multiple writing in MIPHY_PLL_SBR_1, the writing of the
first 0 it's to be sure there is no previous request.
Then we take account new setting by writing 0x02.
And then we make it 0 to make sure there is no other pending requests.
I added comments and macro to be more clear (see the code below).
Hi Kishon,
Do you want a new patch set (v4), or i wait other remarks from you ?
for (val = 0; val < 2; val++) {
writeb_relaxed(val, miphy_phy->base + MIPHY_CONF);
/* Validate Step component */
writeb_relaxed(0x69, miphy_phy->base + MIPHY_PLL_SBR_3);
writeb_relaxed(0x21, miphy_phy->base + MIPHY_PLL_SBR_4);
/* Validate Period component */
writeb_relaxed(0x3c, miphy_phy->base + MIPHY_PLL_SBR_2);
writeb_relaxed(0x21, miphy_phy->base + MIPHY_PLL_SBR_4);
/* Clear any previous request */
writeb_relaxed(0x00, miphy_phy->base + MIPHY_PLL_SBR_1);
/* requests the PLL to take in account new parameters */
writeb_relaxed(SET_NEW_CHANGE, base + MIPHY_PLL_SBR_1);
/* To be sure there is no other pending requests */
writeb_relaxed(0x00, miphy_phy->base + MIPHY_PLL_SBR_1);
}
Best regards
Gabriel.
On 29 September 2014 21:19, <Valdis.Kletnieks@xxxxxx> wrote:
> On Fri, 26 Sep 2014 10:54:15 +0200, Gabriel FERNANDEZ said:
>> SSC is the technique of modulating the operating frequency of a signal
>> slightly to spread its radiated emissions over a range of frequencies.
>> This reduction in the maximum emission for a given frequency helps meet
>> radiated emission requirements.
>> These settings are applicable for PCIE with Internal clock.
>
>> + writeb_relaxed(0x69, miphy_phy->base + MIPHY_PLL_SBR_3);
>> + writeb_relaxed(0x21, miphy_phy->base + MIPHY_PLL_SBR_4);
>> + writeb_relaxed(0x3c, miphy_phy->base + MIPHY_PLL_SBR_2);
>> + writeb_relaxed(0x21, miphy_phy->base + MIPHY_PLL_SBR_4);
>> + writeb_relaxed(0x00, miphy_phy->base + MIPHY_PLL_SBR_1);
>> + writeb_relaxed(0x02, miphy_phy->base + MIPHY_PLL_SBR_1);
>> + writeb_relaxed(0x00, miphy_phy->base + MIPHY_PLL_SBR_1);
>
> I'd feel a lot better about all these magic numbers (and the triple write
> to SBR_1) if the Changelog or something referenced where they came from....
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
