On 30 August 2013 11:13, Viresh Kumar viresh.kumar@linaro.org wrote:
On 29 August 2013 19:12, Santosh Shilimkar santosh.shilimkar@ti.com wrote:
On Thursday 29 August 2013 09:39 AM, Viresh Kumar wrote:
On 29 August 2013 19:07, Santosh Shilimkar santosh.shilimkar@ti.com wrote:
On Thursday 29 August 2013 06:02 AM, Viresh Kumar wrote:
cpumask_setall(policy->cpus);
Thats should work I guess. Infact I changed this i downstream kernel a while back but forgot to send a patch. Just see if for some reason above can break UP machine since this driver is used on UP machines as well. Other than that, it should be fine
Atleast I can't see why it will break on UP machines as that should set masks to individual CPUs then...
Okay I will fix that in my longest patchset ever... 230+ patches already :)
WoW !! Looks like you re-wrote most of the CPUFREQ drivers ;-)
I am trying hard to get rid of redundancy of code wherever possible .. :)
I am pasting my patch here for now (don't want to post this separately) as most of the concerned people are following this chain..
I will include it in my patchset and will post it on lists with them (After your Ack, which you will give now :) )
--------x-----------------x--------------
OMG, Gmail fucked up completely... What to do now :)
Earlier (few days back): - We could have selected the plain text mode while sending emails, where emails are corrupted a bit and so patches couldn't be applied. - BUT they are atleast readable as it wasn't fucked up so badly..
But today I noticed that gmail has completely removed tabs from mails (Atleast while pasting code). And so all tabs used for alignments in mail are gone.. And so aren't even readable anymore..
Deselecting plain text mode, sends it in HTML and that's fucked up too.. Mails gets bounced from our lists (and even then gmail has eaten up the tabs from code)...
For now see the attached patch, I will send it using git later..
Btw, this is the link to master patch that you may need during reviews.. https://lkml.org/lkml/2013/8/21/485
From: Viresh Kumar viresh.kumar@linaro.org Date: Wed, 21 Aug 2013 19:05:05 +0530 Subject: [PATCH] cpufreq: omap: use cpufreq_generic_init() routine
Use generic cpufreq_generic_init() routine instead of replicating the same code here.
This also rearranges the code a bit to make it more sensible. Also removes some unnecessary checks.
Signed-off-by: Viresh Kumar viresh.kumar@linaro.org
drivers/cpufreq/omap-cpufreq.c | 41 +++++++++++------------------------------ 1 file changed, 11 insertions(+), 30 deletions(-)
diff --git a/drivers/cpufreq/omap-cpufreq.c b/drivers/cpufreq/omap-cpufreq.c index 09150fc..a94c4ef 100644 --- a/drivers/cpufreq/omap-cpufreq.c +++ b/drivers/cpufreq/omap-cpufreq.c @@ -112,50 +112,31 @@ static inline void freq_table_free(void)
static int omap_cpu_init(struct cpufreq_policy *policy) {
- int result = 0;
int result;
policy->clk = clk_get(NULL, "cpufreq_ck"); if (IS_ERR(policy->clk)) return PTR_ERR(policy->clk);
- if (policy->cpu >= NR_CPUS) {
- result = -EINVAL;
- goto fail_ck;
- }
- if (!freq_table)
- if (!freq_table) { result = opp_init_cpufreq_table(mpu_dev, &freq_table);
- if (result) {
- dev_err(mpu_dev, "%s: cpu%d: failed creating freq table[%d]\n",
- if (result) {
- dev_err(mpu_dev,
- "%s: cpu%d: failed creating freq table[%d]\n", __func__, policy->cpu, result);
- goto fail_ck;
goto fail;
} }
atomic_inc_return(&freq_table_users);
- result = cpufreq_table_validate_and_show(policy, freq_table);
- if (result)
- goto fail_table;
- /*
- On OMAP SMP configuartion, both processors share the voltage
- and clock. So both CPUs needs to be scaled together and hence
- needs software co-ordination. Use cpufreq affected_cpus
- interface to handle this scenario. Additional is_smp() check
- is to keep SMP_ON_UP build working.
- */
- if (is_smp())
- cpumask_setall(policy->cpus);
- /* FIXME: what's the actual transition time? */
- policy->cpuinfo.transition_latency = 300 * 1000;
- return 0;
- result = cpufreq_generic_init(policy, freq_table, 300 * 1000);
- if (!result)
- return 0;
-fail_table: freq_table_free(); -fail_ck: +fail: clk_put(policy->clk); return result; }