It used to be that I could choose Sleep when shutting down,
then wake the computer by hitting any key. (I've always set
the keyboard to be able to wake while disabling that option
for the mouse.)
With my latest build I have options in the BIOS. One is
something I've never seen before: A setting to wake from
either OS or BIOS. If I enable S3 and set to wake from the
BIOS, the computer can be woken via keyboard, but it can
also be started from cold by touching the keyboard! There
are also other options of how to wake it, such as USB. But
since I have a PS2 trackball I'm able to have it wake only
via PS2 keyboard.
Does that mean it's always in some version of sleep if
I select wake via BIOS?
If instead I choose to wake from the OS, the keyboard no
longer responds, and the BIOS choices disappear, but I can
wake it by pressing the blinking
power button. Perhaps I should mention that I've built this
computer reusing an older case, so the power button is
probably 10 years old or so. But the motherboard connections
seem to still be the same.
All of this is not really a problem, but I'm trying to understand
my options and how this "modern" design works under the surface.
On Sat, 8 Jun 2024 09:15:59 -0400, Newyana2 <newyana@invalid.nospam>
wrote:
It used to be that I could choose Sleep when shutting down,
then wake the computer by hitting any key. (I've always set
the keyboard to be able to wake while disabling that option
for the mouse.)
With my latest build I have options in the BIOS. One is
something I've never seen before: A setting to wake from
either OS or BIOS. If I enable S3 and set to wake from the
BIOS, the computer can be woken via keyboard, but it can
also be started from cold by touching the keyboard! There
are also other options of how to wake it, such as USB. But
since I have a PS2 trackball I'm able to have it wake only
via PS2 keyboard.
Does that mean it's always in some version of sleep if
I select wake via BIOS?
If instead I choose to wake from the OS, the keyboard no
longer responds, and the BIOS choices disappear, but I can
wake it by pressing the blinking
power button. Perhaps I should mention that I've built this
computer reusing an older case, so the power button is
probably 10 years old or so. But the motherboard connections
seem to still be the same.
All of this is not really a problem, but I'm trying to understand
my options and how this "modern" design works under the surface.
I have a separate switch to turn off the monitor overnight. I can't
get the MSoft menu to work.
It used to be that I could choose Sleep when shutting down,
then wake the computer by hitting any key. (I've always set
the keyboard to be able to wake while disabling that option
for the mouse.)
With my latest build I have options in the BIOS. One is
something I've never seen before: A setting to wake from
either OS or BIOS. If I enable S3 and set to wake from the
BIOS, the computer can be woken via keyboard, but it can
also be started from cold by touching the keyboard! There
are also other options of how to wake it, such as USB. But
since I have a PS2 trackball I'm able to have it wake only
via PS2 keyboard.
Does that mean it's always in some version of sleep if
I select wake via BIOS?
If instead I choose to wake from the OS, the keyboard no
longer responds, and the BIOS choices disappear, but I can
wake it by pressing the blinking
power button. Perhaps I should mention that I've built this
computer reusing an older case, so the power button is
probably 10 years old or so. But the motherboard connections
seem to still be the same.
All of this is not really a problem, but I'm trying to understand
my options and how this "modern" design works under the surface.
https://forum-en.msi.com/index.php?attachments/1646496155729-jpeg.157510/
To me, only the BIOS choice makes sense. If the PC is in S4, then
we need the BIOS to handle PME (power management event, any event
raising the power state of a peripheral interface -- originally
used for Wake On LAN). The Windows setting in Device Manager,
to "allow this device to wake the computer", the OS sends the info
to the BIOS, and the BIOS is the last one to ensure the USB port
is ready to wake when a PME happens.
I have a separate switch to turn off the monitor overnight. I can't
get the MSoft menu to work.
On 6/9/2024 5:54 PM, Peter Jason wrote:
I have a separate switch to turn off the monitor overnight. I can't
get the MSoft menu to work.
You mean you can't get your computer to go into sleep
mode at all but you leave the computer on constantly?
That's unfortunate.
On 6/9/2024 10:30 PM, Paul wrote:
https://forum-en.msi.com/index.php?attachments/1646496155729-jpeg.157510/Yes. That's it. I have a B760M-P. I've actually had Asus boards
for many years. This is the first MSI board in a long time. It seems
to be fine, generally. My last Asus was also fine.
To me, only the BIOS choice makes sense. If the PC is in S4, then
we need the BIOS to handle PME (power management event, any event
raising the power state of a peripheral interface -- originally
used for Wake On LAN). The Windows setting in Device Manager,
to "allow this device to wake the computer", the OS sends the info
to the BIOS, and the BIOS is the last one to ensure the USB port
is ready to wake when a PME happens.
When I choose BIOS I have a choice to sleep in S3 and
wake up via USB device or PS2. The trouble is that the
trackball is USB. I don't want that to be able to cause a wake
because any vibration can set it off.
I realize now that I mixed things up in my initial description. The keyboard is PS2. Not the trackball. I chose the PS2 option so
that I could be sure only pressing a key would wake it up. (I was
surprised to find that the motherboard has both PS2 ports.)
Now only the power button wakes it up. My hesitation with
BIOS option is that I don't like the idea of the BIOS standing by.
I don't want wake on LAN or any other wake option other than
the keyboard. But I suppose that if I'm waking via the OS then
the BIOS must still be standing by, no?
Mainly I just don't want any security risks from remote, and
I don't want the mouse/trackball to be able to wake it up.
The Powercfg utility is a power house of options. A very useful utility
On 6/9/2024 10:54 PM, Newyana2 wrote:
On 6/9/2024 10:30 PM, Paul wrote:
https://forum-en.msi.com/index.php?attachments/1646496155729-jpeg.157510/ >>>Yes. That's it. I have a B760M-P. I've actually had Asus boards
for many years. This is the first MSI board in a long time. It seems
to be fine, generally. My last Asus was also fine.
To me, only the BIOS choice makes sense. If the PC is in S4, then
we need the BIOS to handle PME (power management event, any event
raising the power state of a peripheral interface -- originally
used for Wake On LAN). The Windows setting in Device Manager,
to "allow this device to wake the computer", the OS sends the info
to the BIOS, and the BIOS is the last one to ensure the USB port
is ready to wake when a PME happens.
When I choose BIOS I have a choice to sleep in S3 and
wake up via USB device or PS2. The trouble is that the
trackball is USB. I don't want that to be able to cause a wake
because any vibration can set it off.
I realize now that I mixed things up in my initial description. The
keyboard is PS2. Not the trackball. I chose the PS2 option so
that I could be sure only pressing a key would wake it up. (I was
surprised to find that the motherboard has both PS2 ports.)
Now only the power button wakes it up. My hesitation with
BIOS option is that I don't like the idea of the BIOS standing by.
I don't want wake on LAN or any other wake option other than
the keyboard. But I suppose that if I'm waking via the OS then
the BIOS must still be standing by, no?
Mainly I just don't want any security risks from remote, and
I don't want the mouse/trackball to be able to wake it up.
The BIOS arms the PCH (Southbridge) for PME events. The BIOS
then normally goes to sleep.
It's hardware, a logic block, that is watching the store. There
is no program activity to speak of.
Newer hardware could have some whizzy way of waking for short
intervals. Nothing I have here does that, but my gear isn't all
that fancy. I assume even S0IC devices have low power states
when the user isn't using the machine. S0IC allows being in S3,
and then waking up for 500 usec to deal with, maybe Wifi or something. Perhaps it can keep connections alive, while the machine "nominally"
sleeps. A kind of "fitful sleep", a lot of tossing and turning.
When the BIOS wakes up, it passes control to the OS as soon as
initialization is complete. The OS also has initialization after
sleep, as the drivers need to prime the registers on the
chips that did not have power. That is regarded as a kind of
"warm start", since you were sleeping, the code for the driver
is already loaded, but the registers are not the way you like them.
On 6/10/2024 1:43 AM, Paul wrote:
On 6/9/2024 10:54 PM, Newyana2 wrote:
On 6/9/2024 10:30 PM, Paul wrote:
https://forum-en.msi.com/index.php?attachments/1646496155729-jpeg.157510/ >>>>Yes. That's it. I have a B760M-P. I've actually had Asus boards
for many years. This is the first MSI board in a long time. It seems
to be fine, generally. My last Asus was also fine.
To me, only the BIOS choice makes sense. If the PC is in S4, then
we need the BIOS to handle PME (power management event, any event
raising the power state of a peripheral interface -- originally
used for Wake On LAN). The Windows setting in Device Manager,
to "allow this device to wake the computer", the OS sends the info
to the BIOS, and the BIOS is the last one to ensure the USB port
is ready to wake when a PME happens.
When I choose BIOS I have a choice to sleep in S3 and
wake up via USB device or PS2. The trouble is that the
trackball is USB. I don't want that to be able to cause a wake
because any vibration can set it off.
I realize now that I mixed things up in my initial description. The
keyboard is PS2. Not the trackball. I chose the PS2 option so
that I could be sure only pressing a key would wake it up. (I was
surprised to find that the motherboard has both PS2 ports.)
Now only the power button wakes it up. My hesitation with
BIOS option is that I don't like the idea of the BIOS standing by.
I don't want wake on LAN or any other wake option other than
the keyboard. But I suppose that if I'm waking via the OS then
the BIOS must still be standing by, no?
Mainly I just don't want any security risks from remote, and
I don't want the mouse/trackball to be able to wake it up.
The BIOS arms the PCH (Southbridge) for PME events. The BIOS
then normally goes to sleep.
It's hardware, a logic block, that is watching the store. There
is no program activity to speak of.
Newer hardware could have some whizzy way of waking for short
intervals. Nothing I have here does that, but my gear isn't all
that fancy. I assume even S0IC devices have low power states
when the user isn't using the machine. S0IC allows being in S3,
and then waking up for 500 usec to deal with, maybe Wifi or something.
Perhaps it can keep connections alive, while the machine "nominally"
sleeps. A kind of "fitful sleep", a lot of tossing and turning.
When the BIOS wakes up, it passes control to the OS as soon as
initialization is complete. The OS also has initialization after
sleep, as the drivers need to prime the registers on the
chips that did not have power. That is regarded as a kind of
"warm start", since you were sleeping, the code for the driver
is already loaded, but the registers are not the way you like them.
So where does all that leave security? Would it be accurate to
say that neither PC nor BIOS wake is insecure, but that there's
always a remote possibility that some kind of hack coming down
the wire or the airwaves could wake them up if ethernet and/or
wifi are connected?
On 6/10/2024 7:50 AM, Newyana2 wrote:
On 6/10/2024 1:43 AM, Paul wrote:
So where does all that leave security? Would it be accurate to
say that neither PC nor BIOS wake is insecure, but that there's
always a remote possibility that some kind of hack coming down
the wire or the airwaves could wake them up if ethernet and/or
wifi are connected?
But wouldn't that attack surface always be present ?
On 6/10/2024 8:37 AM, Paul wrote:
On 6/10/2024 7:50 AM, Newyana2 wrote:
On 6/10/2024 1:43 AM, Paul wrote:
That's what I'm not clear about. I can disable wake-on-LAN,So where does all that leave security? Would it be accurate to
say that neither PC nor BIOS wake is insecure, but that there's
always a remote possibility that some kind of hack coming down
the wire or the airwaves could wake them up if ethernet and/or
wifi are connected?
But wouldn't that attack surface always be present ?
but the possibility of it implies that an outside signal can start
the computer, possibly even when I consider it to be off and
not asleep.
I don't understand how that works or what form a hack
might take. I suppose the easiest is just to leave it unplugged,
which is what I've generally done in the past. All my computers
are running hardwired to Internet and the main one I usually
unplug when not in use.
On 6/9/2024 5:54 PM, Peter Jason wrote:
On Sat, 8 Jun 2024 09:15:59 -0400, Newyana2 <newyana@invalid.nospam>
wrote:
It used to be that I could choose Sleep when shutting down,
then wake the computer by hitting any key. (I've always set
the keyboard to be able to wake while disabling that option
for the mouse.)
With my latest build I have options in the BIOS. One is
something I've never seen before: A setting to wake from
either OS or BIOS. If I enable S3 and set to wake from the
BIOS, the computer can be woken via keyboard, but it can
also be started from cold by touching the keyboard! There
are also other options of how to wake it, such as USB. But
since I have a PS2 trackball I'm able to have it wake only
via PS2 keyboard.
Does that mean it's always in some version of sleep if
I select wake via BIOS?
If instead I choose to wake from the OS, the keyboard no
longer responds, and the BIOS choices disappear, but I can
wake it by pressing the blinking
power button. Perhaps I should mention that I've built this
computer reusing an older case, so the power button is
probably 10 years old or so. But the motherboard connections
seem to still be the same.
All of this is not really a problem, but I'm trying to understand
my options and how this "modern" design works under the surface.
I have a separate switch to turn off the monitor overnight. I can't
get the MSoft menu to work.
Better to not rely on a mechanical switch which can fail.
A Shortcut using the free
NirSoft\multimonitortool-x64\MultiMonitorTool.exe can turn off one or
two monitors.
On 6/9/2024 5:54 PM, Peter Jason wrote:
I have a separate switch to turn off the monitor overnight. I can't
get the MSoft menu to work.
You mean you can't get your computer to go into sleep
mode at all but you leave the computer on constantly?
That's unfortunate.
I never use the UEFI/BIOS wake/sleep option, just Windows Power Options.
Inactivity configurations
- On desktop (Win10 Pro)
Display off at 15 min, SSD at 20 min, Computer sleep at 30 min
- On laptop (Win11 Pro)
For Battery settings Display 10 min, SSD Never, Computer sleep 20 min
For Plug in settings Display 15 min, SSD Never, Computer sleep 30 min
On 6/10/2024 11:34 AM, Newyana2 wrote:
On 6/10/2024 8:37 AM, Paul wrote:
On 6/10/2024 7:50 AM, Newyana2 wrote:That's what I'm not clear about. I can disable wake-on-LAN,
On 6/10/2024 1:43 AM, Paul wrote:
So where does all that leave security? Would it be accurate to
say that neither PC nor BIOS wake is insecure, but that there's
always a remote possibility that some kind of hack coming down
the wire or the airwaves could wake them up if ethernet and/or
wifi are connected?
But wouldn't that attack surface always be present ?
but the possibility of it implies that an outside signal can start
the computer, possibly even when I consider it to be off and
not asleep.
I don't understand how that works or what form a hack
might take. I suppose the easiest is just to leave it unplugged,
which is what I've generally done in the past. All my computers
are running hardwired to Internet and the main one I usually
unplug when not in use.
In Device Manager (right-click Start menu Win10, to find the item),
then look at the Properties of the NIC.
In the Power Management tab of the NIC properties, it says
Allow this device to wake the computer
If that box is unticked, then the Ethernet LED on the switch/router
should be un-lit, when the computer does a shutdown, and the Ethernet
is then not supposed to be listening for WOL. The Allow this device to wake the computer, should affect the PME path from the NIC to the rest of
the machine. But the BIOS also interprets this state request, as a
request to power off the core of the NIC (saving one watt). There's no
point having the NIC core running, if it is no longer possible to wake
the PC. That's regular ACPI, the S0ic behavior could be different.
| Sysop: | Keyop |
|---|---|
| Location: | Huddersfield, West Yorkshire, UK |
| Users: | 431 |
| Nodes: | 16 (2 / 14) |
| Uptime: | 156:24:42 |
| Calls: | 9,070 |
| Files: | 13,402 |
| Messages: | 6,020,577 |
