cpufreq regression due to dceff5c "cpufreq: fix serialization issues with freq change notifiers"
One of the test cases I run (not often enough, hence I just noticed this late) is the following, to make sure our cpufreq driver is present, and adjusting the CPU clock as expected:
# echo 1000000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq ; echo 216000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq
which prints the following when it works:
1000000
216000
In next-20130909, that doesn't work, but instead prints:
216000
216000
I've also seen the following during git bisect:
608000
608000
... and perhaps other unexpected sets of values.
This is a regression. I found that if I revert dceff5c "cpufreq: fix serialization issues with freq change notifiers", then the test case works as expected.
Any idea what's up?
Quoted 19 lines · Stephen Warren
One of the test cases I run (not often enough, hence I just noticed this late) is the following, to make sure our cpufreq driver is present, and adjusting the CPU clock as expected:
# echo 1000000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq ; echo 216000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freqwhich prints the following when it works:
1000000
216000In next-20130909, that doesn't work, but instead prints:
216000
216000
Quoted 6 lines
I've also seen the following during git bisect:
608000
608000... and perhaps other unexpected sets of values.
Can you please test the current linux-next? I've reverted the commit that dceff5c attempted to fix, so the output should be correct now.
Quoted 5 lines
This is a regression. I found that if I revert dceff5c "cpufreq: fix serialization issues with freq change notifiers", then the test case works as expected.
Any idea what's up?
Commit dceff5c overlooked some places in which transition_ongoing should have been taken care of.
Thanks,
Rafael
Signature · 2 lines
I speak only for myself. Rafael J. Wysocki, Intel Open Source Technology Center.
Quoted 22 lines · Rafael J. Wysocki
Quoted 21 lines · Stephen Warren
One of the test cases I run (not often enough, hence I just noticed this late) is the following, to make sure our cpufreq driver is present, and adjusting the CPU clock as expected:
# echo 1000000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq ; echo 216000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freqwhich prints the following when it works:
1000000
216000In next-20130909, that doesn't work, but instead prints:
216000
216000I've also seen the following during git bisect:
Quoted 7 lines
Quoted 4 lines
608000
608000... and perhaps other unexpected sets of values.
Can you please test the current linux-next? I've reverted the commit that dceff5c attempted to fix, so the output should be correct now.
Yes, indeed next-20130911 works fine. Thanks, and sorry for the noise.
Quoted 26 lines · Stephen Warren
Quoted 25 lines · Rafael J. Wysocki
Quoted 24 lines · Stephen Warren
One of the test cases I run (not often enough, hence I just noticed this late) is the following, to make sure our cpufreq driver is present, and adjusting the CPU clock as expected:
# echo 1000000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq ; echo 216000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freqwhich prints the following when it works:
1000000
216000In next-20130909, that doesn't work, but instead prints:
216000
216000I've also seen the following during git bisect:
608000
608000
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
... and perhaps other unexpected sets of values.
Can you please test the current linux-next? I've reverted the commit that dceff5c attempted to fix, so the output should be correct now.
Yes, indeed next-20130911 works fine. Thanks, and sorry for the noise.
No problem and thanks for the confirmation!
Quoted 26 lines · Stephen Warren
Quoted 25 lines · Rafael J. Wysocki
Quoted 24 lines · Stephen Warren
One of the test cases I run (not often enough, hence I just noticed this late) is the following, to make sure our cpufreq driver is present, and adjusting the CPU clock as expected:
# echo 1000000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq ; echo 216000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_min_freq ; sleep 1 ; cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freqwhich prints the following when it works:
1000000
216000In next-20130909, that doesn't work, but instead prints:
216000
216000I've also seen the following during git bisect:
608000
608000
Quoted 6 lines
Quoted 4 lines
Quoted 1 line
... and perhaps other unexpected sets of values.
Can you please test the current linux-next? I've reverted the commit that dceff5c attempted to fix, so the output should be correct now.
Yes, indeed next-20130911 works fine. Thanks, and sorry for the noise.
For the record this is where this problem is getting discussed:
Quoted 3 lines · Viresh Kumar
For the record this is where this problem is getting discussed:
Can you help us a bit here? That patch was reverted for now but we are planning to get it back and I have few patches which must help us here...
If you can try them on Tegra, it would be great!!
All are present here:
Quoted 5 lines · Viresh Kumar
Quoted 3 lines · Viresh Kumar
For the record this is where this problem is getting discussed:
Quoted 7 lines
Can you help us a bit here? That patch was reverted for now but we are planning to get it back and I have few patches which must help us here...
If you can try them on Tegra, it would be great!!
All are present here:
I'm not sure how relevant it is anymore given the discussion in response to the patches you posted, but I tested anyway. Let me know if there's anything else you need me to test; I'm a little confused with all the overlapping/alternate cpufreq patches flying by recently. git branches are easiest for me.
==========
Branch cpufreq-suspend-fix
Commit 8f777cc "cpufreq: use correct values of cpus in __cpufreq_remove_dev_finish()"
That works fine; I can both suspend/resume multiple times, and force cpufreq frequency, and everything appears to work.
==========
Branch cpufreq-transition-ongoing
Commit 42851e3 "cpufreq: fix notification serialization issues"
Forcing cpufreq appears to work fine for me in this branch. However, suspend/resume has problems:
1) On the first resume, I see sysfs complaining:
Quoted 62 lines
[ 27.215037] Enabling non-boot CPUs ... [ 27.226919] CPU1: Booted secondary processor [ 27.232832] ------------[ cut here ]------------ [ 27.238026] WARNING: CPU: 0 PID: 810 at fs/sysfs/dir.c:530 sysfs_add_one+0x80/0xa0() [ 27.246566] sysfs: cannot create duplicate filename '/devices/system/cpu/cpu0/cpufreq/stats' [ 27.255679] Modules linked in: [ 27.259222] CPU: 0 PID: 810 Comm: bash Not tainted 3.11.0-02563-gf7baf38 #144 [ 27.267097] [<c00156b4>] (unwind_backtrace+0x0/0xf8) from [<c00116d8>] (show_stack+0x10/0x14) [ 27.276418] [<c00116d8>] (show_stack+0x10/0x14) from [<c050da30>] (dump_stack+0x80/0xc4) [ 27.285201] [<c050da30>] (dump_stack+0x80/0xc4) from [<c0024474>] (warn_slowpath_common+0x64/0x88) [ 27.294952] [<c0024474>] (warn_slowpath_common+0x64/0x88) from [<c002452c>] (warn_slowpath_fmt+0x30/0x40) [ 27.305359] [<c002452c>] (warn_slowpath_fmt+0x30/0x40) from [<c011e8b4>] (sysfs_add_one+0x80/0xa0) [ 27.315109] [<c011e8b4>] (sysfs_add_one+0x80/0xa0) from [<c011eab8>] (create_dir+0x5c/0xbc) [ 27.324202] [<c011eab8>] (create_dir+0x5c/0xbc) from [<c011ed60>] (sysfs_create_subdir+0x24/0x2c) [ 27.333858] [<c011ed60>] (sysfs_create_subdir+0x24/0x2c) from [<c0120590>] (internal_create_group+0x50/0x25c) [ 27.344648] [<c0120590>] (internal_create_group+0x50/0x25c) from [<c034688c>] (cpufreq_stat_notifier_policy+0x98/0x28c) [ 27.356402] [<c034688c>] (cpufreq_stat_notifier_policy+0x98/0x28c) from [<c0042f48>] (notifier_call_chain+0x44/0x84) [ 27.367855] [<c0042f48>] (notifier_call_chain+0x44/0x84) from [<c004311c>] (__blocking_notifier_call_chain+0x48/0x60) [ 27.379424] [<c004311c>] (__blocking_notifier_call_chain+0x48/0x60) from [<c004314c>] (blocking_notifier_call_chain+0x18/0x20) [ 27.391830] [<c004314c>] (blocking_notifier_call_chain+0x18/0x20) from [<c0344fa8>] (__cpufreq_set_policy+0xac/0x1c0) [ 27.403355] [<c0344fa8>] (__cpufreq_set_policy+0xac/0x1c0) from [<c03450ec>] (cpufreq_init_policy+0x30/0x70) [ 27.414047] [<c03450ec>] (cpufreq_init_policy+0x30/0x70) from [<c0346228>] (__cpufreq_add_dev.isra.27+0x494/0x608) [ 27.425291] [<c0346228>] (__cpufreq_add_dev.isra.27+0x494/0x608) from [<c0346414>] (cpufreq_cpu_callback+0x70/0x8c) [ 27.436643] [<c0346414>] (cpufreq_cpu_callback+0x70/0x8c) from [<c0042f48>] (notifier_call_chain+0x44/0x84) [ 27.447251] [<c0042f48>] (notifier_call_chain+0x44/0x84) from [<c0024590>] (__cpu_notify+0x28/0x44) [ 27.457095] [<c0024590>] (__cpu_notify+0x28/0x44) from [<c0024818>] (_cpu_up+0xf0/0x140) [ 27.465882] [<c0024818>] (_cpu_up+0xf0/0x140) from [<c05099d8>] (enable_nonboot_cpus+0x68/0xb0) [ 27.475376] [<c05099d8>] (enable_nonboot_cpus+0x68/0xb0) from [<c005d7d8>] (suspend_devices_and_enter+0x19c/0x2f8) [ 27.486652] [<c005d7d8>] (suspend_devices_and_enter+0x19c/0x2f8) from [<c005daa8>] (pm_suspend+0x174/0x1e8) [ 27.497255] [<c005daa8>] (pm_suspend+0x174/0x1e8) from [<c005c928>] (state_store+0x6c/0xbc) [ 27.506361] [<c005c928>] (state_store+0x6c/0xbc) from [<c01e4940>] (kobj_attr_store+0x14/0x20) [ 27.515697] [<c01e4940>] (kobj_attr_store+0x14/0x20) from [<c011d3f0>] (sysfs_write_file+0x168/0x198) [ 27.525731] [<c011d3f0>] (sysfs_write_file+0x168/0x198) from [<c00cbed0>] (vfs_write+0xb0/0x188) [ 27.535301] [<c00cbed0>] (vfs_write+0xb0/0x188) from [<c00cc280>] (SyS_write+0x3c/0x70) [ 27.544029] [<c00cc280>] (SyS_write+0x3c/0x70) from [<c000e540>] (ret_fast_syscall+0x0/0x30) [ 27.553277] ---[ end trace f27199d855bd2598 ]--- [ 27.553883] ------------[ cut here ]------------ [ 27.553925] WARNING: CPU: 1 PID: 810 at fs/sysfs/dir.c:530 sysfs_add_one+0x80/0xa0() [ 27.553935] sysfs: cannot create duplicate filename '/devices/system/cpu/cpu0/cpufreq/stats' [ 27.553948] Modules linked in: [ 27.553974] CPU: 1 PID: 810 Comm: bash Tainted: G W 3.11.0-02563-gf7baf38 #144 [ 27.554042] [<c00156b4>] (unwind_backtrace+0x0/0xf8) from [<c00116d8>] (show_stack+0x10/0x14) [ 27.554087] [<c00116d8>] (show_stack+0x10/0x14) from [<c050da30>] (dump_stack+0x80/0xc4) [ 27.554126] [<c050da30>] (dump_stack+0x80/0xc4) from [<c0024474>] (warn_slowpath_common+0x64/0x88) [ 27.554160] [<c0024474>] (warn_slowpath_common+0x64/0x88) from [<c002452c>] (warn_slowpath_fmt+0x30/0x40) [ 27.554197] [<c002452c>] (warn_slowpath_fmt+0x30/0x40) from [<c011e8b4>] (sysfs_add_one+0x80/0xa0) [ 27.554233] [<c011e8b4>] (sysfs_add_one+0x80/0xa0) from [<c011eab8>] (create_dir+0x5c/0xbc) [ 27.554267] [<c011eab8>] (create_dir+0x5c/0xbc) from [<c011ed60>] (sysfs_create_subdir+0x24/0x2c) [ 27.554308] [<c011ed60>] (sysfs_create_subdir+0x24/0x2c) from [<c0120590>] (internal_create_group+0x50/0x25c) [ 27.554350] [<c0120590>] (internal_create_group+0x50/0x25c) from [<c034688c>] (cpufreq_stat_notifier_policy+0x98/0x28c) [ 27.554401] [<c034688c>] (cpufreq_stat_notifier_policy+0x98/0x28c) from [<c0042f48>] (notifier_call_chain+0x44/0x84) [ 27.554445] [<c0042f48>] (notifier_call_chain+0x44/0x84) from [<c004311c>] (__blocking_notifier_call_chain+0x48/0x60) [ 27.554488] [<c004311c>] (__blocking_notifier_call_chain+0x48/0x60) from [<c004314c>] (blocking_notifier_call_chain+0x18/0x20) [ 27.554526] [<c004314c>] (blocking_notifier_call_chain+0x18/0x20) from [<c0344fa8>] (__cpufreq_set_policy+0xac/0x1c0) [ 27.554561] [<c0344fa8>] (__cpufreq_set_policy+0xac/0x1c0) from [<c0345a80>] (cpufreq_update_policy+0x90/0xec) [ 27.554596] [<c0345a80>] (cpufreq_update_policy+0x90/0xec) from [<c034641c>] (cpufreq_cpu_callback+0x78/0x8c) [ 27.554632] [<c034641c>] (cpufreq_cpu_callback+0x78/0x8c) from [<c0042f48>] (notifier_call_chain+0x44/0x84) [ 27.554669] [<c0042f48>] (notifier_call_chain+0x44/0x84) from [<c0024590>] (__cpu_notify+0x28/0x44) [ 27.554701] [<c0024590>] (__cpu_notify+0x28/0x44) from [<c0024818>] (_cpu_up+0xf0/0x140) [ 27.554745] [<c0024818>] (_cpu_up+0xf0/0x140) from [<c05099d8>] (enable_nonboot_cpus+0x68/0xb0) [ 27.554797] [<c05099d8>] (enable_nonboot_cpus+0x68/0xb0) from [<c005d7d8>] (suspend_devices_and_enter+0x19c/0x2f8) [ 27.554840] [<c005d7d8>] (suspend_devices_and_enter+0x19c/0x2f8) from [<c005daa8>] (pm_suspend+0x174/0x1e8)
Quoted 8 lines
[ 27.554879] [<c005daa8>] (pm_suspend+0x174/0x1e8) from [<c005c928>] (state_store+0x6c/0xbc) [ 27.554920] [<c005c928>] (state_store+0x6c/0xbc) from [<c01e4940>] (kobj_attr_store+0x14/0x20) [ 27.554956] [<c01e4940>] (kobj_attr_store+0x14/0x20) from [<c011d3f0>] (sysfs_write_file+0x168/0x198) [ 27.554992] [<c011d3f0>] (sysfs_write_file+0x168/0x198) from [<c00cbed0>] (vfs_write+0xb0/0x188) [ 27.555027] [<c00cbed0>] (vfs_write+0xb0/0x188) from [<c00cc280>] (SyS_write+0x3c/0x70) [ 27.555065] [<c00cc280>] (SyS_write+0x3c/0x70) from [<c000e540>] (ret_fast_syscall+0x0/0x30) [ 27.555079] ---[ end trace f27199d855bd2599 ]--- [ 27.555151] CPU1 is up
2) On the second suspend, I see a crash; I think the same one we've been discussing before:
Quoted 57 lines
[ 47.983000] Disabling non-boot CPUs ... [ 48.003941] Unable to handle kernel NULL pointer dereference at virtual address 00000000 [ 48.029770] pgd = ed9cc000 [ 48.054576] [00000000] *pgd=2d978831, *pte=00000000, *ppte=00000000 [ 48.079572] Internal error: Oops: 817 [#1] PREEMPT SMP ARM [ 48.103709] Modules linked in: [ 48.125565] CPU: 0 PID: 810 Comm: bash Tainted: G W 3.11.0-02563-gf7baf38 #144 [ 48.153195] task: ed444a80 ti: edb0a000 task.ti: edb0a000 [ 48.178409] PC is at cpufreq_stat_notifier_policy+0x21c/0x28c [ 48.204310] LR is at notifier_call_chain+0x44/0x84 [ 48.229463] pc : [<c0346a10>] lr : [<c0042f48>] psr: 60000113 [ 48.229463] sp : edb0bdb0 ip : 00000000 fp : 000aa2d8 [ 48.282667] r10: c0fec048 r9 : c073cf40 r8 : ee281c80 [ 48.309108] r7 : 00000004 r6 : 00000000 r5 : ee281c80 r4 : 00000000 [ 48.336891] r3 : 00000000 r2 : 008bc000 r1 : 00000000 r0 : 00000001 [ 48.364323] Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user [ 48.392331] Control: 10c5387d Table: 2d9cc04a DAC: 00000015 [ 48.419065] Process bash (pid: 810, stack limit = 0xedb0a238) [ 48.445723] Stack: (0xedb0bdb0 to 0xedb0c000) [ 48.470712] bda0: ed90da10 00000000 c073ccb4 ffffffff [ 48.499809] bdc0: 00000000 00000000 00000004 ee281c80 c073cf40 c0fec048 000aa2d8 c0042f48 [ 48.529033] bde0: c076599c ffffffff ee281c80 00000004 00000001 c004311c 00000000 00200200 [ 48.558145] be00: ed444a80 ee281c80 00000001 00000000 c07c9068 c004314c 00000000 c034476c [ 48.587155] be20: c07310ac c0344a7c 00000015 00000001 c07cb9e4 00000015 00000001 ee08ec40 [ 48.615942] be40: c073d804 c0346428 c03463a4 ffffffde edb0be88 c0042f48 00000001 c073c414 [ 48.644546] be60: c073ccb4 00000010 00000001 c0024590 edb0be88 c073c414 c073ccb4 c0509778 [ 48.672930] be80: 000aa2d8 c050c40c 00000021 edb0bea4 00000010 00000001 00000001 c073c414 [ 48.701175] bea0: c073ccb4 00000000 c0780080 c0024a84 c07837ac 00000000 00000003 c07837ac [ 48.729214] bec0: c003ede0 c005d79c 000aa2d8 c050c40c c065d4d0 edb0beec 00000003 edb0beec [ 48.757174] bee0: 00000000 00000003 c0516a80 00000003 ee33e000 ee08ec40 00000004 c005daa8 [ 48.785030] bf00: 0000006d 00000003 c06349e8 c005c928 00000004 00000004 ed86fac0 ed86fad8 [ 48.812686] bf20: edb0bf80 c05310c4 ee0e5868 c01e4940 00000004 c011d3f0 ed4e6140 00000004 [ 48.840184] bf40: 000ac408 edb0bf80 00000000 edb0a000 00000004 c00cbed0 edab6500 00000001 [ 48.867552] bf60: 0000000a 00000000 00000000 ed4e6140 000ac408 00000000 00000004 c00cc280 [ 48.894723] bf80: 00000000 00000000 ee1474c0 b6f61a78 00000004 000ac408 00000004 c000e6c4 [ 48.921639] bfa0: 00000000 c000e540 b6f61a78 00000004 00000001 000ac408 00000004 00000000 [ 48.948312] bfc0: b6f61a78 00000004 000ac408 00000004 be95e95c 000a6094 00000000 000aa2d8 [ 48.974906] bfe0: 00000000 be95e8dc b6ed1b77 b6f0825c 40070010 00000001 2e7f6821 2e7f6c21 [ 49.001489] [<c0346a10>] (cpufreq_stat_notifier_policy+0x21c/0x28c) from [<c0042f48>] (notifier_call_chain+0x44/0x84) [ 49.031019] [<c0042f48>] (notifier_call_chain+0x44/0x84) from [<c004311c>] (__blocking_notifier_call_chain+0x48/0x60) [ 49.060673] [<c004311c>] (__blocking_notifier_call_chain+0x48/0x60) from [<c004314c>] (blocking_notifier_call_chain+0x18/0x20) [ 49.110276] [<c004314c>] (blocking_notifier_call_chain+0x18/0x20) from [<c0344a7c>] (__cpufreq_remove_dev_prepare.isra.11+0x140/0x27c) [ 49.162054] [<c0344a7c>] (__cpufreq_remove_dev_prepare.isra.11+0x140/0x27c) from [<c0346428>] (cpufreq_cpu_callback+0x84/0x8c) [ 49.214698] [<c0346428>] (cpufreq_cpu_callback+0x84/0x8c) from [<c0042f48>] (notifier_call_chain+0x44/0x84) [ 49.245939] [<c0042f48>] (notifier_call_chain+0x44/0x84) from [<c0024590>] (__cpu_notify+0x28/0x44) [ 49.276638] [<c0024590>] (__cpu_notify+0x28/0x44) from [<c0509778>] (_cpu_down+0x80/0x238) [ 49.306512] [<c0509778>] (_cpu_down+0x80/0x238) from [<c0024a84>] (disable_nonboot_cpus+0x68/0xe8) [ 49.337171] [<c0024a84>] (disable_nonboot_cpus+0x68/0xe8) from [<c005d79c>] (suspend_devices_and_enter+0x160/0x2f8) [ 49.369662] [<c005d79c>] (suspend_devices_and_enter+0x160/0x2f8) from [<c005daa8>] (pm_suspend+0x174/0x1e8) [ 49.401642] [<c005daa8>] (pm_suspend+0x174/0x1e8) from [<c005c928>] (state_store+0x6c/0xbc) [ 49.432373] [<c005c928>] (state_store+0x6c/0xbc) from [<c01e4940>] (kobj_attr_store+0x14/0x20) [ 49.463408] [<c01e4940>] (kobj_attr_store+0x14/0x20) from [<c011d3f0>] (sysfs_write_file+0x168/0x198) [ 49.495146] [<c011d3f0>] (sysfs_write_file+0x168/0x198) from [<c00cbed0>] (vfs_write+0xb0/0x188) [ 49.526605] [<c00cbed0>] (vfs_write+0xb0/0x188) from [<c00cc280>] (SyS_write+0x3c/0x70) [ 49.557445] [<c00cc280>] (SyS_write+0x3c/0x70) from [<c000e540>] (ret_fast_syscall+0x0/0x30) [ 49.588767] Code: e5950010 e7922100 e783c002 e595300c (e5813000) [ 49.618011] ---[ end trace f27199d855bd259b ]---
Quoted 5 lines · Stephen Warren
I'm not sure how relevant it is anymore given the discussion in response to the patches you posted, but I tested anyway. Let me know if there's anything else you need me to test; I'm a little confused with all the overlapping/alternate cpufreq patches flying by recently. git branches are easiest for me.
I believe everybody (hopefully leaving me) is confused :) ... There are so many patches around same set of problems going on.. And most of them came from me or Srivatsa, so I hope I know which ones to pick :)
And yes, thanks a lot for testing these..
Quoted 7 lines
==========
Branch cpufreq-suspend-fix
Commit 8f777cc "cpufreq: use correct values of cpus in__cpufreq_remove_dev_finish()"That works fine; I can both suspend/resume multiple times, and force cpufreq frequency, and everything appears to work.
Good.. So, it proves the theory that Myself and Srivatsa have discussed.. For single cluster platform my patch is fine and for multicluster ones Srivatsa's patch is required..
Quoted 8 lines
==========
Branch cpufreq-transition-ongoing
Commit 42851e3 "cpufreq: fix notification serialization issues"Forcing cpufreq appears to work fine for me in this branch. However, suspend/resume has problems:
1) On the first resume, I see sysfs complaining:
Okay.. that's expected... Both sets are required for a perfect solution :) These were two different bugs and so separate branches.