I’m really looking forward for coreboot firmware on Framework devices. 
6 Likes
We can but dream
I complained about this before in one of the Framework 16 threads, however really what bothers me is when there’s either functionality waiting on a bios update or when there’s a bug that’s affecting me that I’m waiting for an update for. Otherwise, honestly I could really care less.
The bios updates contain obviously updates to the bios itself, but also stuff like PD firmware, processor microcode updates, etc… But unless there’s some bug that someone wants/needs fixed, really what I think most people are looking for are microcode updates. However, microcode updates can also be loaded early in the boot process before the operating system itself. I believe this happens automatically in Windows, and can be enabled in Linux if not enabled by default. You’re not really missing anything with it not being included in the bios, it’s just getting loaded a bit later in the boot process than if it was.
I was annoyed with the delay in the update for the Framework 16 because I was waiting for the release of the dual M.2 expansion bay board, which required a bios update to work properly. It’s doubly annoying that they “leaked” the availability of said board in an LTT video 6+ months before they were able to release a bios update, knowing that they likely had stacks of them available for months but weren’t able to sell them because of needing this update. Had it not leaked and otherwise kept quiet about the release of said board until it was ready, I wouldn’t have known it was coming and therefore wouldn’t have cared about the update.
In the Framework 16 thread I compared Framework to the likes of Dell, HP, and Lenovo, who push out bios updates sometimes near monthly for a good chunk of their lineup, generally with descriptions like “security updates”. But these are likely automated with little to no human interaction and probably only update microcode month to month. Once you start looking are smaller manufacturers, bios updates come few and far between, sometimes never getting an update at all. For instance, I own a Gigabyte Aero 16 12th gen (XE5), which only ever got a single bios update all the way back in September of 2022. I have an Asus Zenbook Space Edition that got updates about every 6 months, but the most recent update is over a year old now. I have a few Minisforum devices that, if they received any updates at all, likely won’t be updated at all now since newer versions of said devices have been released. Similarly with GPD. Then you have Tongfang and Clevo; good luck getting any updates for those at all, and if you do you’re downloading updates from Google Drive or MEGA via some random forum somewhere that someone posted a link to claiming it’s an update for a particular device. Maybe they got it from a legit source, maybe they didn’t, who knows.
All that said, it would be nice for Framework to provide updates. It’s rather annoying when they miss a deadline they just kind of disappear behind the curtain until they randomly have an update a number of weeks/months later.
1 Like
I doubt the inclusion of microcode is causing delays to Framework releasing BIOS updates.
I didn’t say that it was; I was arguing that you don’t really need a bios update unless there’s some bug, and that churning them out for microcode updates really doesn’t do much since the operating system is just going to load the same or newer microcode anyway.
when is the next bios release?
Well, I also don’t care when or if at all a manufacturer releases new features for a BIOS, but I care a lot about security fixes, and it’s simply unacceptable that it takes that long, and even worse, missing their own deadlines without even an update for when the update can be expected.
Yes, some security issues can maybe fixed with microcode updates, but not all of them, for example LogoFAIL. I’m also not sure if all microcode updates are available as separate packages, or if AMD only provides some updates as part of a BIOS update. So in these cases it’s important that a security fix can be released fast.
It might be annoying, that you need to wait for a leaked feature, but there can also be other reasons why a BIOS update should be released which is more than just annoying. For example when there is broken functionality, like when the 3.03 update broke compatibility with a lot of chargers (and I think some people received their laptops with already this version installed, so it was broken from the start), and people weren’t able anymore to charge their laptops. This also took them almost half a year to fix with the next version, while it was advertised that people can also use other chargers and don’t need to buy the framework charger. And that wasn’t just a “leak” and a laptop you can’t charge is a bit more than just “annoying”.
I also don’t really care about what other manufacturers do. And it doesn’t matter that some are even worse and don’t release any bios updates. And I rather have automated monthly updates with security updates only, than no updates at all.
And I see a laptop with a known security vulnerability or a laptop which can’t charge as just broken, and it should be fixed. And I don’t know why new BIOS versions are taking that long. It looks like this one is hold up by a new feature which is unrelated to the other changes. It might be possible to even install this beta and just disable the broken feature. But I don’t think people should need to install a held back beta as the only option to fix their laptop. If the feature isn’t ready yet, why can’t they just release a version without that feature and then I also don’t care anymore how long it takes them to finish that feature I was never promised (don’t get me wrong, I also like it when they add new features, and it’s a nice feature and I hope it will be released at some day … but it shouldn’t hold back other important updates).
Maybe they should look more into automation, so they can release small security-fixes faster and more automated and develop new features in a separate branch and test the feature in a beta that is unrelated to other important fixes. Framework is already doing a lot of things well, but this is clearly a place where there is still a lot of room for improvements (as this isn’t the first time I’m experiencing this, and friend already had similar experiences with an intel framework laptop). And it would already help if they would keep us updated what the current state is and when we can expect which fixes, instead of just ghosting their customers.
Surely BIOS updates that are just microcode updates are rare anyway. I think maybe the last one I did on my desktop PC was just microcode, but all the others before that one fixed bugs and/or added features, often including fixing security problems.
So, I guess this is no longer the target (as it is now over 2 weeks ago). Any update on the fix? I was happy to test the BETA, but it does come with an expectation that fixes for issues that are found will be provided in a reasonable timeframe. As it stands, beta testers are left with a Bios with defects.
10 Likes
agree with you on reasonable timeframe for the bugfix. it will be much better if Framework can provide us with some kind of schedule/roadmap. But to be fair… at least on my laptop. the 3.06 BETA after disabling the battery extender does not have any negative effect. I mostly have it plugged in, but it still runs 5 hours unplugged (Pop! without any optimization)
1 Like
Same here, and, as far as I know, the only negative effect it can have is that an LED keeps flashing between two colours, so not a showstopper.
I’m not sure this is true for AMD-based laptops. Would be great if it was. But this thread, with a reference to a post on the Phoronix forum, suggests that most AMD microcode releases are only happening through the AGESA updates, for this class of AMD chipsets.
1 Like
That was true back in 2023 however consumer AMD chips are included in the linux-firmware repository now.
For instance, here’s the family, model, and stepping for the 7840U:
Jan 19 04:10:39 archlinux kernel: smpboot: CPU0: AMD Ryzen 7 7840U w/ Radeon 780M Graphics (family: 0x19, model: 0x74, stepping: 0x1)
And then from the list of microcode patches I linked to:
Family=0x19 Model=0x74 Stepping=0x01: Patch=0x0a704107 Length=5568 bytes
And then if we check the boot logs to see if the microcode was updated:
Jan 19 04:10:40 archlinux kernel: microcode: Current revision: 0x0a704107
Jan 19 04:10:40 archlinux kernel: microcode: Updated early from: 0x0a704104
So according to this, the version in the bios was 0x0a704104 and was updated to 0x0a704107 during boot, and therefore we are, in fact, getting microcode updates via linux-firmware. I assume that there’s a similar mechanism in windows.
6 Likes
I can confirm that this worked for me as well. Switching the iGPU to gaming mode in the BIOS allowed me to install theatest drivers without issue.
don’t know if related to this BIOS, but notice having this issue occasionally.
1 Like
I know there were the holidays to contend with, but the inability to get BIOS updates out in a timely manner is the most frustrating part of my Framework experience. Admittedly the first time around was far worse because we went 6 months without Framework even acknowledging that there was an issue causing resets, let alone getting the BIOS update out… but now we’re at nearly two months since this beta was pulled without a resolution in sight.
I believe Framework primarily outsources firmware development, I’m not sure you’re getting your money’s worth. 
9 Likes
Well, one month late now and no update on the post. Whilst I want to help with Beta testing, this will be the last time I do so with a BIOS release.
6 Likes
I don’t think anyone can really BETA test the BIOS any more, because it does not let you rollback if there is a problem. A test without a rollback is not a well thought out plan.
9 Likes
If 3.07 beta is a fix for 3.06 beta… I will hesitate, but might just install it to see what happen.
4 Likes
Name a more iconic duo than Framework and late BIOS updates.
4 Likes