| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Fri, 16 Nov 2018 21:55:51 +0100
From: Martin Blumenstingl <martin.blumenstingl@...glemail.com>
To: linux-amlogic@...ts.infradead.org,
Neil Armstrong <narmstrong@...libre.com>, jbrunet@...libre.com
Cc: mturquette@...libre.com, sboyd@...nel.org,
linux-clk@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 0/3] Meson8b: add the CPU clock post-dividers
On Fri, Nov 16, 2018 at 9:53 PM Martin Blumenstingl
<martin.blumenstingl@...glemail.com> wrote:
>
> This is the successor to my previous series "meson8b: add the CPU_DIV16
> clock for the ARM TWD" from [0]. I decided to not send this as v2 of
> the original series because the PERIPH clock is not the CPU_DIV16 clock.
> It's not clear whether a CPU_DIV16 clock exists.
>
> With this series we get all the CPU_CLK post-dividers as listed in the
> public S805 datasheet [1] on pages 31 and 32:
> - ABP
> - PERIPH (used as input for the ARM global timer and ARM TWD timer)
> - AXI
> - L2 DRAM
>
> Each of these clocks has a register called "..._CLK_DIS" which is
> documented as a "just in case" bit:
> "Set to 1 to manually disable the [...] clock when changing the mux
> selection. Typically this bit is set to 0 since the clock muxes can
> switch without glitches."
> Since we're not supposed to touch that register we're setting
> CLK_IS_CRITICAL for these clocks in the driver.
>
> The result of this is that we can use the PERIPH clock which clocks
> the ARM TWD timer. I will send a separate series to add the TWD timer.
>
>
> [0] http://lists.infradead.org/pipermail/linux-amlogic/2018-July/007890.html
> [1] https://dn.odroid.com/S805/Datasheet/S805_Datasheet%20V0.8%2020150126.pdf
>
> Martin Blumenstingl (3):
> dt-bindings: clock: meson8b: export the CPU post dividers
> clk: meson: meson8b: rename cpu_div2/cpu_div3 to
> cpu_in_div2/cpu_in_div3
> clk: meson: meson8b: add the CPU clock post divider clocks
I ran various tests for the PERIPH clock by enabling the TWD timer
without specifying a clock. In that case the TWD driver will try to
auto-detect the PERIPH clock rate. my results confirm that this
implementation is correct:
case #1: EC-100 with it's default boot settings:
[ 0.000252] smp_twd: clock not found -2
[ 0.000736] Calibrating local timer... 197.86MHz.
# cat /sys/kernel/debug/clk/clk_summary
enable prepare protect
duty
clock count count count rate
accuracy phase cycle
---------------------------------------------------------------------------------------------
...
sys_pll_dco 1 1 0 1584000000
0 0 50000
sys_pll 1 1 0 792000000
0 0 50000
cpu_in_sel 1 1 0 792000000
0 0 50000
cpu_scale_out_sel 1 1 0 792000000
0 0 50000
cpu_clk 4 4 0 792000000
0 0 50000
...
cpu_clk_div4 2 2 0 198000000
0 0 50000
axi_clk_sel 1 1 0
198000000 0 0 50000
axi_clk_dis 1 1 0
198000000 0 0 50000
periph_clk_sel 1 1 0
198000000 0 0 50000
periph_clk_dis 1 1 0
198000000 0 0 50000
...
case #2: switch the parent of the cpu_in_sel clock to xtal:
m8b_m201_1G#md.l 0xC110419C 1
c110419c: 000000b1 ....
m8b_m201_1G#mw.l 0xC110419C 0xb0 1
m8b_m201_1G#md.l 0xC110419C
c110419c: 000000b0 ....
[ 0.007005] smp_twd: clock not found -2
[ 0.022680] Calibrating local timer... 5.98MHz.
# cat /sys/kernel/debug/clk/clk_summary
enable prepare protect
duty
clock count count count rate
accuracy phase cycle
---------------------------------------------------------------------------------------------
xtal 4 4 0 24000000
0 0 50000
...
cpu_in_sel 1 1 0 24000000
0 0 50000
cpu_scale_out_sel 1 1 0 24000000
0 0 50000
cpu_clk 4 4 0 24000000
0 0 50000
...
cpu_clk_div4 2 2 0 6000000
0 0 50000
axi_clk_sel 1 1 0 6000000
0 0 50000
axi_clk_dis 1 1 0 6000000
0 0 50000
periph_clk_sel 1 1 0 6000000
0 0 50000
periph_clk_dis 1 1 0
6000000 0 0 50000
...
case #3: set cpu_scale_div to cpu_in_sel/6 and use cpu_scale_div as
parent of cpu_scale_out_sel
m8b_m201_1G#mw.l 0xC110419C 0x30 1
m8b_m201_1G#md.l 0xC110415C 1
c110415c: 800016a2 ....
m8b_m201_1G#mw.l 0xC110415C 0x802016A2 1
m8b_m201_1G#mw.l 0xC110419C 0xbd 1
[ 0.001317] smp_twd: clock not found -2
[ 0.003907] Calibrating local timer... 32.93MHz.
# cat /sys/kernel/debug/clk/clk_summary
enable prepare protect
duty
clock count count count rate
accuracy phase cycle
---------------------------------------------------------------------------------------------
...
sys_pll_dco 1 1 0 1584000000
0 0 50000
sys_pll 1 1 0 792000000
0 0 50000
cpu_in_sel 1 1 0 792000000
0 0 50000
cpu_scale_div 1 1 0 132000000
0 0 50000
cpu_scale_out_sel 1 1 0
132000000 0 0 50000
cpu_clk 4 4 0 132000000
0 0 50000
...
cpu_clk_div4 2 2 0
33000000 0 0 50000
axi_clk_sel 1 1 0
33000000 0 0 50000
axi_clk_dis 1 1 0
33000000 0 0 50000
periph_clk_sel 1 1 0
33000000 0 0 50000
periph_clk_dis 1 1 0
33000000 0 0 50000
...
case #4: set cpu_scale_div to cpu_in_sel/4 and use cpu_scale_div as
parent of cpu_scale_out_sel
m8b_m201_1G#mw.l 0xC110419C 0x30 1
m8b_m201_1G#md.l 0xC110415C 1
c110415c: 800016a2 ....
m8b_m201_1G#mw.l 0xC110415C 0x801016A2 1
m8b_m201_1G#mw.l 0xC110419C 0xbd 1
[ 0.001317] smp_twd: clock not found -2
[ 0.002636] Calibrating local timer... 49.42MHz.
# cat /sys/kernel/debug/clk/clk_summary
enable prepare protect
duty
clock count count count rate
accuracy phase cycle
---------------------------------------------------------------------------------------------
...
sys_pll_dco 1 1 0 1584000000
0 0 50000
sys_pll 1 1 0 792000000
0 0 50000
cpu_in_sel 1 1 0 792000000
0 0 50000
cpu_scale_div 1 1 0 198000000
0 0 50000
cpu_scale_out_sel 1 1 0
198000000 0 0 50000
cpu_clk 4 4 0 198000000
0 0 50000
...
cpu_clk_div4 2 2 0
49500000 0 0 50000
axi_clk_sel 1 1 0
49500000 0 0 50000
axi_clk_dis 1 1 0
49500000 0 0 50000
periph_clk_sel 1 1 0
49500000 0 0 50000
periph_clk_dis 1 1 0
49500000 0 0 50000
...
case #5: set cpu_scale_div to cpu_in_sel/12 and use cpu_scale_div as
parent of cpu_scale_out_sel
m8b_m201_1G#mw.l 0xC110419C 0x30 1
m8b_m201_1G#md.l 0xC110415C 1
c110415c: 800016a2 ....
m8b_m201_1G#mw.l 0xC110415C 0x805016A2 1
m8b_m201_1G#mw.l 0xC110419C 0xbd 1
[ 0.001317] smp_twd: clock not found -2
[ 0.007728] Calibrating local timer... 16.44MHz.
# cat /sys/kernel/debug/clk/clk_summary
enable prepare protect
duty
clock count count count rate
accuracy phase cycle
---------------------------------------------------------------------------------------------
...
sys_pll_dco 1 1 0 1584000000
0 0 50000
sys_pll 1 1 0 792000000
0 0 50000
cpu_in_sel 1 1 0 792000000
0 0 50000
cpu_scale_div 1 1 0 66000000
0 0 50000
cpu_scale_out_sel 1 1 0
66000000 0 0 50000
cpu_clk 4 4 0 66000000
0 0 50000
...
cpu_clk_div4 2 2 0
16500000 0 0 50000
axi_clk_sel 1 1 0
16500000 0 0 50000
axi_clk_dis 1 1 0
16500000 0 0 50000
periph_clk_sel 1 1 0
16500000 0 0 50000
periph_clk_dis 1 1 0
16500000 0 0 50000
...
case #6: use cpu_div2 as parent of cpu_scale_out_sel
m8b_m201_1G#mw.l 0xC110419C 0x30 1
m8b_m201_1G#mw.l 0xC110419C 0x34 1
m8b_m201_1G#mw.l 0xC110419C 0xb5 1
[ 0.000534] smp_twd: clock not found -2
[ 0.001487] Calibrating local timer... 98.91MHz.
# cat /sys/kernel/debug/clk/clk_summary
enable prepare protect
duty
clock count count count rate
accuracy phase cycle
---------------------------------------------------------------------------------------------
...
sys_pll_dco 1 1 0 1584000000
0 0 50000
sys_pll 1 1 0 792000000
0 0 50000
cpu_in_sel 1 1 0 792000000
0 0 50000
...
cpu_div2 1 1 0 396000000
0 0 50000
cpu_scale_out_sel 1 1 0
396000000 0 0 50000
cpu_clk 4 4 0 396000000
0 0 50000
...
cpu_clk_div4 2 2 0
99000000 0 0 50000
axi_clk_sel 1 1 0
99000000 0 0 50000
axi_clk_dis 1 1 0
99000000 0 0 50000
periph_clk_sel 1 1 0
99000000 0 0 50000
periph_clk_dis 1 1 0
99000000 0 0 50000
...
case #7: use cpu_div3 as parent of cpu_scale_out_sel
m8b_m201_1G#mw.l 0xC110419C 0x30 1
m8b_m201_1G#mw.l 0xC110419C 0x38 1
m8b_m201_1G#mw.l 0xC110419C 0xb9 1
test skipped, the system is not stable with this clock as parent
Powered by blists - more mailing lists