Laptop basics: i7-2620M cores AMD graphics 7.45GiB RAM’s 953GiB total SSD’s if(window.hljsLoader && !document.currentScript.parentNode.hasAttribute('data-s9e-livepreview-onupdate')) { window.hljsLoader.highlightBlocks(document.currentScript.parentNode); } FreeBSD-13.0-CURRENT-amd64-20200806-r363935-memstick.img : Failed (for both CloverEFI and legacy BIOS modes) Below those are untested , FreeBSD-13.0-CURRENT-amd64-20200806-r363935-disc1.iso FreeBSD-13.0-CURRENT-amd64-20200730-r363681-memstick.img FreeBSD-13.0-CURRENT-amd64-20200723-r363439-memstick.img FreeBSD-13.0-CURRENT-amd64-20200716-r363236-memstick.img FreeBSD-13.0-CURRENT-amd64-20200730-r363681-disc1.iso FreeBSD-13.0-CURRENT-amd64-20200723-r363439-disc1.iso FreeBSD-13.0-CURRENT-amd64-20200716-r363236-disc1.iso if(window.hljsLoader && !document.currentScript.parentNode.hasAttribute('data-s9e-livepreview-onupdate')) { window.hljsLoader.highlightBlocks(document.currentScript.parentNode); } https://forums.freebsd.org/threads/sony-made-vaio-vpccb17fg-issues.76547

hd_scania At the loader screen, choose 6. (for boot options), then switch both Safe mode and Verbose to on. Is this a new problem, or did FreeBSD boot OK on this laptop before?

The discussion continued at https://forums.freebsd.org/threads/sony-made-vaio-vpccb17fg-issues.76547/

aponomarenko Looks like you have things well in hand over there. Thanks, however, for keeping us informed.

Yep, fixed by changing the USB stick and adding kern.cam.ada.write_cache=0 boot option.

And i need to also get the wifi chip back to working connections My chip is codenamed ath0 on FreeBSD and to be assigned as wlan0 And which commands can do this? I took a break from BSD for about 0.5 years Now i need to back to BSD thru FreeBSD, then when done, i need to take some snaps to show my FreeBSD desktops on a new hardware☺️

The wireless issue was gone too, but the XFCE shell was NO mice available with the touchpad disabled , i have TWO USB and wireless mice but BOTH were NOT working And finally, the Plasma shell was blunk in black, so i’m yet to need to also test out Enlightenment and Lumina i’ve also installed on FreeBSD, alongside Plasma and XFCE

Does /etc/rc.conf contain moused_enable="YES" ? Have you made any changes to the Xorg config files?

rvp Yes and moused_type has been set to auto too, but ahead moused_enable=YES And i haven't even touched the Xorg configs at all

And given there were kernel issues which got the system constantly crashed too (and now done by a note on the FreeBSD docs on kernel debugging), finally was likely getting moused crashed too

[13-CURRENT] Sony-made VAIO VPCCB17FG issues

hd_scania

The battery LED on the FreeBSD side was ALWAYS gone red, i.e. it reports the battery there was CRITICAL
On the Linux side, on the other hand, the battery LED is being OFF, i.e. the battery is on a safe level
And pardon? I just haven’t even touched any CloverEFI configs, and on rEFInd no power settings are provided, but why still boot FreeBSD ahead CloverEFI? This hugely affects that i can NO LONGER boot Linux, and as to unhide the Linux loaders, it requires CloverEFI reinstalled

rvp

hd_scania The battery LED on the FreeBSD side was ALWAYS gone red, i.e. it reports the battery there was CRITICAL
On the Linux side, on the other hand, the battery LED is being OFF, i.e. the battery is on a safe level

I think you're mistaken--check this again.

Battery LED colour and blinking, and also charging, on your laptop, is controlled by the BIOS+EC (Embedded Controller). It is not under control of the OS. If battery status is CRITICAL on FreeBSD, then it will be the same on Linux also. (And, we've already seen that the battery values match.)

hd_scania I just haven’t even touched any CloverEFI configs,...

If you haven't made any ACPI-related overrides, then you can keep CloverEFI/rEFInd.

My verdict is that your battery is to blame, and not FreeBSD, for the mysterious shutdowns. Keep it on AC power always and I think FreeBSD should be stable.

And, if you use the correct binary graphics drivers (drm-kmod or xf86-video-...) with 12.1-RELEASE, then Plasma should work too (again, on AC power becaue your battery is flaky). If you run, -STABLE or -CURRENT, then, install the graphics drivers from source via ports.

There is one FreeBSD-related bug here, though, which could have been resposible for your WiFi not working earlier:

WARNING: attempt to domain_add(bluetooth) after domainfinalize()
WARNING: attempt to domain_add(netgraph) after domainfinalize()

If you see these 2 lines at boot-time, then networking won't work. I see these messages on FreeBSD-12.1-RELEASE when I boot into it directly after using Linux--ie. without a shutdown in between. The "fix" I use is to shutdown my laptop, and wait 10 seconds before booting into any (other) OS.

hd_scania

Flaky? UNLIKELY thx? And yes i’ve checked, the battery has a lock, but was NOT tightened
So it could be concluded that, the issue was caused by the BAD BATTERY CONNECTIONS
You got it SOLVED up to here☺️

My verdict is that your battery is to be blamed, instead of software, for the mysterious shutdowns. Keep it on AC power always and I think FreeBSD should be stable.
And, if you use the correct binary graphics drivers (drm-kmod or xf86-video-...) with 12.1-RELEASE, then Plasma should work too (again, on AC power becaue your battery is flaky).

hd_scania

PS i’ve had an low end laptop too (an Sony-made VAIO, too, Japan rocks😎😺), and just checked its battery at my first time on it, it’s been always properly sealed

hd_scania

The system was crashed again, but this time the battery is no longer loose, up to now, i’m going to expect critical kernel / ACPI bugs
On BOTH FreeBSD and Linux sides the battery LED was remained OFF, i.e. the battery levels are NO LONGER CRITICAL
Here, we need to overcome some FreeBSD-exclusively kernel / ACPI bugs, which is NEVER a Linux case
I.e. now is a PURELY SOFTWARE issue solely on the FreeBSD userland

rvp

hd_scania On BOTH FreeBSD and Linux sides the battery LED was remained OFF,

If the battery LED is off, it indicates that the laptop is off AC power. On AC power it should be green when charged and amber when charging.

Check the AC plug and clean the internal battery cotacts (use isopropyl alcohol).
Post output of the batt.sh script when running on/off AC power on the console.
Does the battery LED light up when on AC power?

hd_scania

Unfortunately yes, the AC WAS loose, now the battery LED is being red
I need to stay at Linux until that LED is changed to green
...
A few minutes later, the LED was backed to OFF again somewhy
Does is suggest that the issue is actually come from the transformer?
It feels like that, my laptop CAN’T be charged, but the transformer’s LED was ALWAYS GREEN
But the laptop battery was always being RED OR OFF, but whose lock is now NO LONGER LOOSE

rvp

hd_scania A few minutes later, the LED was backed to OFF again somewhy
Does is suggest that the issue is actually come from the transformer?

Open the laptop, then check the AC socket on the motherboard. It might be loose because the solder has weakened/come off.

hd_scania but the transformer’s LED was ALWAYS GREEN

That means the AC power brick is OK. Forget about that. Note only the laptop's battery indicator LED.

hd_scania

Given the laptop's FINAL BIOS upgrade was released at 2011, the AC female was expected to be loose, due to already being aged for at least a decade
I need to send it to the original reseller for further actions, he IS expected to repair as an external-market PC reseller

rvp

hd_scania the AC female was expected to be loose, due to already being aged for at least a decade

So, the AC barrel socket was loose?
You can easily solder it yourself (or pay an electronics engineering student to do it for you) after you buy a replacement socket.

It is an interesting laptop, though. I don't know many laptops with 2 graphics cards in it:

vgapci1@pci0:0:2:0: class=0x030000 card=0x9080104d chip=0x01268086 rev=0x09 hdr=0x00
    vendor     = 'Intel Corporation'
    device     = '2nd Generation Core Processor Family Integrated Graphics Controller'
    class      = display
    subclass   = VGA
    ...
vgapci0@pci0:1:0:0: class=0x030000 card=0x9080104d chip=0x67411002 rev=0x00 hdr=0x00
    vendor     = 'Advanced Micro Devices, Inc. [AMD/ATI]'
    device     = 'Whistler [Radeon HD 6630M/6650M/6750M/7670M/7690M]'
    class      = display
    subclass   = VGA

hd_scania

Yeah i paid for someone in 10GBP to do for me
AC barrel socket? Unsure wtf barrel socket was, but the AC female is still being loose

rvp

hd_scania the AC female is still being loose

Same thing. Also called a receptacle, jack, ...

hd_scania

On my BOTH BSD and Linux sides, under a same battery, but too differently seen to the BOTH systems on a completely same battery
On no-systemd Linux: The battery state normal
On FreeBSD: The battery NEVER above -1% but NEVER rechargeable
Even though the transform brick is renewed, but still, the FreeBSD side has always abnormally perceived a totally wrong battery dataset

hd_scania

Then here, UNLIKELY hardware issues at all, but instead, there’ve been some fatal ACPI or kernel bugs too wrongly reading the battery data

rvp

hd_scania Then here, UNLIKELY hardware issues at all, but instead, there’ve been some fatal ACPI or kernel bugs too wrongly reading the battery data

Not exactly a FreeBSD kernel bug (see post #32 where we determined this already).

The main problem here is that your battery is providing wrong (voltage) values. The kernel takes that bogus voltage, does a small computation on it, then looks at the result, and says: "Huh?! this doesn't look right", and then removes that battery from the system. Linux, probably just skips using that voltage, but keeps the battery online. I'll have to check and see.

hd_scania And i was asked to add the line hw.x2apic_enable=0 onto the kernel command lines
But does this affect my battery UNREADABLE to the system (only on FreeBSD, but NOT on Linux)?

Highly unlikely. You probably don't need that line in /boot/loader.conf. Just remove it.

For now:

Ignore the battery stats on FreeBSD. Look at the HW/firmware-controlled battery LED, instead.
Open the laptop up and take a photo of the specs. (voltage, mAh, ...) printed on the battery pack. I'll need it to decide how to "fix" your little battery problem.

BTW: while your laptop is opened, clean the dust out from your CPU fan/heatsink. Re-apply thermal paste if you have to. 85 C is too hot for an idle system. (Or, were you compiling GCC?)

And, to get Intel CPU temp. data in hwstat add this line to /boot/loader.conf:

coretemp_load="YES"

rvp

Run the batt.sh script in post #41.

hd_scania

rvp
https://mega.nz/file/eYhBlazS#BsPWWMIMZS28wk2nbxpQO2iXX6kqRAjux9FKX8ZJV5s

hd_scania

And i was asked to add the line hw.x2apic_enable=0 onto the kernel command lines
But does this affect my battery UNREADABLE to the system (only on FreeBSD, but NOT on Linux)?

hd_scania

rvp
But on the other hands,

Without the line hw.x2apic_enable=0 STALL the system from BEFORE the init scripts
But with that line CRASHES the kernel too
But by having added coretemp on the line kldload_list in my rc.conf (or the same to be loaded in /boot/loader.conf), and the first line PERM REMOVED, will just be working fine?

Highly unlikely. You probably don't need that line in /boot/loader.conf. Just remove it.

hd_scania

rvp Open the laptop up and take a photo of the specs. (voltage, mAh, ...) printed on the battery pack. I'll need it to decide how to "fix" your little battery problem.

hd_scania

Yes i’ve just tested a few mins ago, without the line hw.x2apic_enable=0 DOESN’T work, from your coretemp kernel mods to be suggested instead, does this suggest that, my CPU overheat crashes the system?

rvp

hd_scania without the line hw.x2apic_enable=0 DOESN’T work,

Really? I thought it was the other line kern.cam.ada.write_cache=0 (post #5) which fixed that problem. Well, if that option is also needed, keep it.

hd_scania does this suggest that, my CPU overheat crashes the system?

Are you saying now that the system crashes even on A/C power, or on a (mostly) full battery (ie. green battery LED)? I thought this problem was taken care of...

It could be heat-related, I don't know. But, the temperature is much too high for an idle, or mostly idle, system. Running powerd might help. Run these commands as root:

sysrc powerd_enable=YES
sysrc powerd_flags="-a adp -b min"

coretemp_load="YES" in /boot/loader.conf is for reporting the CPU temperature on Intel Core CPUs.

And, which FreeBSD version (RELEASE, STABLE, CURRENT) do you want the fix to be applied to?

« Previous Page Next Page »