You can dramatically increase your laptop battery life by using these two simple tricks...

[Ottavio Caruso]

I've been a Thinkpad fanatic for almost 20 years, but I never realised how silly I was by not disabling these two features.

So:

- Boot into your laptop BIOS/UEFI setup.

- Disable "USB always on" (Your BIOS setup utility may use a different name);

- Disable "Intel Fast-Boot" (If not an Intel CPU, it may be under a different name).

- Save settings and reboot.

Et voila, your old, crappy, near-dead battery will be instantly rejuvenated.

Do this from the BIOS setup, not from your OS power management software.

[Colin Bignell]

That subject line reads like the start of a spam email message.

[Ottavio Caruso]

Am 06/07/2023 um 09:41 schrieb Colin Bignell:

Thank you.

[alan_m]

It seems that laptop and phone manufacturers are recommending only charging the battery to 70 or 85% to increase its working life.

Both my laptop and phone include a facility to stop charging at these limits.

[Colin Bignell]

My phone will charge to 100%, but if it is charging overnight and I have an alarm set for the morning, it will charge slowly, aiming to get to

100% shortly before the alarm.

[nib]

Although, interestingly, for some of their models Tesla recommend charging to 100% regularly, but on slow home charger, not a fast charger.

nib

[mm0fmf]

The message sounds like bollocks too.

Fastboot off means your PC spends longer in the bios and starts up more devices. Fastboot on skips a huge amount of hardware setup and gets the OS loaded quicker. The OS then loads drivers for whatever it thinks you have fitted.

Likewise modern Linux/Windows will put USB ports into suspend when they are not in use.

So it all sounds like bollocks never mind it's from our new troll "Enrico".

[Ottavio Caruso]

No, you're obviously confusing with a different thing. This is Intel fastboot, that speeds up boot by not letting the laptop shutdown properly.

Not true. I am running Debian Bullseye and you can't disable that behaviour via software. You have to do it from the BIOS.

Salutame a mammeta!

[Andy Burns]

Are you talking about Modern Standby aka "S0 low power idle" vs S3 "sleep" states? if you're using Linux rather than Windows, might explain it ...

[Theo]

'Suspend to RAM' (S3) will keep the memory powered to hold the state of the machine. 'Suspend to disc' aka 'hibernate' (S4) will save the state to disc before de-powering the RAM.

Modern standby is S0ix, which is less power-saving than S3 or S4. Microsoft call it InstantGo:

If you're concerned about losing battery when off and not plugged in, make sure you're doing S4. On Linux that typically means having a swap partition larger than your RAM size, and you may need to actively turn on hibernate.

For me it made a big difference on a rarely used laptop, because sleeping for weeks would inevitably drain the battery by the next time I came to use it.

Theo

[Brian Gaff]

Yes certainly so about power saving as the sleep function uses power all the time but if its stopped in bios windows has no control over it but make sure you always shut down fully once you do this, or it may come back corrupted trying to use sleep mode when the power is cut off. On usbs this is because some people like to charge stuff from they computers and of course this will normally continue when the machine is off, so if you do use this method to charge things, then either stop doing it or leave it alone. Brian

[Ottavio Caruso]

What you just wrote is true but 100% off topic.

If you don't specifically disable USB power on from the BIOS setup, all USB ports will be still draining power, even when the PC is powered off, let alone suspended or hibernated.

And if you don't disable Intel Fast-boot from the BIOS setup, the laptop will go into semi-hibernation, even when powered off.

I have made the comparison myself. Before turning the two features off, my external battery would drain from 45% to 5% in 6 hours, with the laptop switched off and Linux configured into "deep" state on suspend or hibernate.

$ cat /sys/power/mem_sleep s2idle [deep]

See? It's already on [deep], but the CPU will ignore it.

Now it keeps charge when switched off and marginally loses some charge when in suspend, but it's an almost dead battery so it's ok.

[Theo]

If you have USB things plugged in, yes, that's worth doing. You lose the ability to wake the machine from an external keyboard though.

What did you have plugged in when you did these tests?

That's only S2 suspend, which is actually not very deep. The CPU is turned off but the RAM is still powered. Suspend should be S3 and hibernate S4.

However it shouldn't be draining the battery that fast in S2. But this is all controlled by ACPI which is down to the vendor firmware - it may be a problem with your specific laptop.

Theo

[Ottavio Caruso]

No, but the whole USB hub is still powered and still draining.

Intel CPUs are known for ignoring Linux ACPI settings.

[Ottavio Caruso]

Am 07/07/2023 um 13:16 schrieb Ottavio Caruso:

As you can see from the output below, it's a mess.

oc@t440:~$ dmesg |grep ACPI [38005.242285] ACPI: EC: interrupt blocked [38005.257489] ACPI: EC: interrupt unblocked [38006.619062] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [38006.621865] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [38012.274423] ACPI: EC: interrupt blocked [38012.325843] ACPI: PM: Preparing to enter system sleep state S3 [38012.328421] ACPI: EC: event blocked [38012.328422] ACPI: EC: EC stopped [38012.328423] ACPI: PM: Saving platform NVS memory [38012.336515] ACPI: PM: Low-level resume complete [38012.336543] ACPI: EC: EC started [38012.336543] ACPI: PM: Restoring platform NVS memory [38012.342013] ACPI: PM: Waking up from system sleep state S3 [38012.348417] ACPI: EC: interrupt unblocked [38012.388663] ACPI: EC: event unblocked [38014.814170] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [38014.816866] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [38034.249382] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [38034.249389] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [38034.249391] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [38034.249393] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [38034.249394] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [38034.249396] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [38034.249397] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50135.473372] ACPI: EC: interrupt blocked [50135.491355] ACPI: EC: interrupt unblocked [50136.618549] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [50136.621526] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [50143.881796] ACPI: EC: interrupt blocked [50143.937069] ACPI: PM: Preparing to enter system sleep state S3 [50143.939643] ACPI: EC: event blocked [50143.939644] ACPI: EC: EC stopped [50143.939644] ACPI: PM: Saving platform NVS memory [50143.948315] ACPI: PM: Low-level resume complete [50143.948342] ACPI: EC: EC started [50143.948342] ACPI: PM: Restoring platform NVS memory [50143.953798] ACPI: PM: Waking up from system sleep state S3 [50143.960087] ACPI: EC: interrupt unblocked [50144.001600] ACPI: EC: event unblocked [50146.413519] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [50146.416422] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [50158.213063] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50158.213071] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50158.213074] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50158.213076] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50158.213078] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50158.213080] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [50158.213082] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62843.595460] ACPI: EC: interrupt blocked [62843.647124] ACPI: PM: Preparing to enter system sleep state S3 [62843.649829] ACPI: EC: event blocked [62843.649830] ACPI: EC: EC stopped [62843.649831] ACPI: PM: Saving platform NVS memory [62843.661033] ACPI: PM: Low-level resume complete [62843.661060] ACPI: EC: EC started [62843.661061] ACPI: PM: Restoring platform NVS memory [62843.666650] ACPI: PM: Waking up from system sleep state S3 [62843.672886] ACPI: EC: interrupt unblocked [62843.716538] ACPI: EC: event unblocked [62846.187516] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [62846.190049] ata1.00: ACPI cmd f5/00:00:00:00:00:a0(SECURITY FREEZE LOCK) filtered out [62852.378232] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62852.378240] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62852.378243] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62852.378245] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62852.378247] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62852.378249] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001) [62852.378251] ACPI: \: failed to evaluate _DSM bf0212f2-788f-c64d-a5b3-1f738e285ade (0x1001)

[Theo]

What settings do Intel CPUs ignore?

What's a mess exactly? From that list, all I can see is:

Linux has decided not to issue the SECURITY FREEZE LOCK command that ACPI tells it to issue to the SSD, probably because it prevents some kinds of sleeping because you have to hard power down the SSD to wake it.

That looks like a bug in the ACPI tables' extension (DSM) for Windows. They shouldn't matter for Linux (in theory).

Otherwise it looks like a regular sleep/wake cycle. For the record, here's mine (also Lenovo):

[514123.702057] Lockdown: systemd-logind: hibernation is restricted; see man kernel_lockdown.7 [514124.773540] PM: suspend entry (s2idle) [514124.779798] Filesystems sync: 0.006 seconds [514124.906029] Freezing user space processes ... (elapsed 0.003 seconds) done. [514124.909465] OOM killer disabled. [514124.909466] Freezing remaining freezable tasks ... (elapsed 0.001 seconds) done. [514124.911169] printk: Suspending console(s) (use no_console_suspend to debug) [514125.554445] ACPI: EC: interrupt blocked [514128.823163] ACPI: EC: interrupt unblocked [514129.434988] nvme nvme0: 8/0/0 default/read/poll queues [514129.630121] usb 3-1: reset full-speed USB device number 2 using xhci_hcd [514129.818011] OOM killer enabled. [514129.818014] Restarting tasks ... [514129.818515] mei_hdcp 0000:00:16.0-b638ab7e-94e2-4ea2-a552-d1c54b627f04: bound 0000:00:02.0 (ops i915_hdcp_component_ops [i915]) [514129.826452] done. [514129.840545] thermal thermal_zone5: failed to read out thermal zone (-61) [514129.972415] PM: suspend exit

Theo