On 21-07-20, 13:43, Pavel Machek wrote:
On Mon 2020-07-20 17:37:50, Greg Kroah-Hartman wrote:
From: Finley Xiao finley.xiao@rock-chips.com
commit 371a3bc79c11b707d7a1b7a2c938dc3cc042fffb upstream.
The function cpu_power_to_freq is used to find a frequency and set the cooling device to consume at most the power to be converted. For example, if the power to be converted is 80mW, and the em table is as follow. struct em_cap_state table[] = { /* KHz mW */ { 1008000, 36, 0 }, { 1200000, 49, 0 }, { 1296000, 59, 0 }, { 1416000, 72, 0 }, { 1512000, 86, 0 }, }; The target frequency should be 1416000KHz, not 1512000KHz.
Fixes: 349d39dc5739 ("thermal: cpu_cooling: merge frequency and power tables")
Wow, this is completely different from the upstream patch.
Right, I have mentioned this in the patch I sent for stable.
https://lore.kernel.org/lkml/bc3978d0b7472c140e4d87f61138168a2a7b995c.159419...
There the loops goes down, not up. The code does not match the changelog here.
Yes, the order is different in earlier kernels but I would say that the changelog still matches as it doesn't necessarily talks about any ordering here.
--- a/drivers/thermal/cpu_cooling.c +++ b/drivers/thermal/cpu_cooling.c @@ -278,11 +278,11 @@ static u32 cpu_power_to_freq(struct cpuf int i; struct freq_table *freq_table = cpufreq_cdev->freq_table;
- for (i = 1; i <= cpufreq_cdev->max_level; i++)
if (power > freq_table[i].power)
- for (i = 0; i < cpufreq_cdev->max_level; i++)
if (power >= freq_table[i].power) break;
- return freq_table[i - 1].frequency;
- return freq_table[i].frequency;
}
Something is very wrong here, if table is sorted like described in the changelog, it will always break at i==0 or i==1... not working at all in the old or the new version.
As I understand from the other email you sent, this works fine now. Right ?