It Turns Out Windows Unconditionally Reserves The First 1MB Of RAM, Linux Was Just Late To Do So

[indepe]

It's 1MB. I have 64GB in my desktop, 16 in my laptop. My compile machine has 128, and I have servers with a TB of RAM. Why on earth would I risk stability for 1MB?!

Of course you don't want to risk stability in any way. However there could be an warning message, or a utility, that does this memory check just as a warning. Although it would make more sense to have that on Windows. That way a bug report can be made towards the motherboard manufacturer.

[Ronshere]

As far as i can remember MS was reserving the first meg way back in DOS. I am surprised that people are surprised by learning this. Maybe everyone just forgot

[waxhead]

I don't think i have ever seen Win 10 crash.

As for Linux crashing a lot less, that is disingenuous, when Windows crashes you see an obvious sign in the form of a blue screen, Linux just freezes and nothing works other than a hard reboot.

Also, there is a red screen of death that RedHat and Suse will display under certain crash scenarios.

Win10 is certainly not bad at all , but how often do you need to RESTART Windows? I use Win10 at work and others use Win10 also. I hear a lot of complaining from time to time that thing crash / hang etc. but not all of this is Windows fault. Drivers and/or flaky hardware is quite often the problem and not so much the core system anymore.

That being said I have the opposite experience than you. Debian have for me been rock solid and the few times it have appeared to hang I can usually wait a while or log in via ssd from a small netbook that I have on my desk, kill X and have the system restored quite easily. Usually with a decent message in dmesg about what happened. Windows on the other side is not that chatty about things that goes haywire and tends to really leave you without options to recover except a reboot.

[xorbe]

Also, don't map/use the last MB either, because buggy/suboptimal prefetch hardware.

[willmore]

On the desktop, sure. But even when 1 GB desktops became commonplace, there were low memory embedded devices in use. It is only very recently that the smallest embedded x86 devices have enough RAM to not care about 1 MB.

x86 embedded devices with less than 1GB of memory?

[sophisticles]

Eh? Windows has crashed for me quite a bIt in 2021. Just a lot less than LINUX.

I don't think i have ever seen Win 10 crash.

As for Linux crashing a lot less, that is disingenuous, when Windows crashes you see an obvious sign in the form of a blue screen, Linux just freezes and nothing works other than a hard reboot.

Also, there is a red screen of death that RedHat and Suse will display under certain crash scenarios.

[AnAccount]

It's not detectable, that's why they apply this unconditionally in the first place.

Well, that is true that detecting wrongful memory usage by the BIOS is probably quite difficult, and in that case the fix is quite cheap.

[AnAccount]

I don't know what's sadder, that you feel it's better to not have a working computer in the event the BIOS corrupts a small 1mb portion of memory or that 7 people upvoted your comment indicating agreement.

I would think it's much more preferable for the OS to take steps to ensure that the system boots so that an end user can find a fix to the problem.

How would you guys go about fixing a system where the BIOS corrupts the ram if both Windows and Linux refused to boot?

And of course Windows does it better, there's a reason why MS is worth billions of dollars and one of the world's most valuable companies and there are maybe 4 Linux based companies that actually make a profit.

I'm sad that so many people seems to be happy using a BIOS with grave memory issues. If it is so badly written that it is writing outside of its memory, what else have it corrupted?
And I'm also surprised so many people think that an unbootable laptop would get in the hands of any end users? I'm sure there are other bugs in a BIOS during the development process, causing boot failures. Those bugs are obviously show stoppers, and cannot be shipped. All I'm saying, is that a BIOS corrupting memory should not be shipped to end users. Refusing to boot would do that.

[AnAccount]

Maybe the correct way, but not the best user experience. Windows I believe made the correct workaround, and Linux has now followed suit now that the issue has become apparent. It's not always best to go with the "correct" or "right" in my opinion, there's always some nuances, and I think this situation is one of them, especially when there's no standardization.

That would in fact have been a better user experience, because I hardly think a laptop would reach the end user if it couldn't boot an operative system. And if it did, then the laptop maker have a severe QA issue, so I would not recommend using such a laptop anyway.

When Windows quietly added this hack back in the day, they actually caused end users of other OS:es problems since hardware makers are happy with booting Windows. So, the hack have led to bad user experiences. So, refusing to boot would have been the correct thing to do. If you let bad programmers get away with bad code, they are going to continue writing bad code...

[neozeed]

I suppose it would possible to run a check after one hour of operation and test that 1 MB of memory, to identify those systems messing with memory that they shouldn't mess with (even if there may be false negatives after a single test of 1 hour).

It's 1MB. I have 64GB in my desktop, 16 in my laptop. My compile machine has 128, and I have servers with a TB of RAM. Why on earth would I risk stability for 1MB?!