Page MenuHomeFreeBSD

Update platform docs for arm
Needs ReviewPublic

Authored by jmg on May 24 2020, 8:12 PM.
Tags
None
Referenced Files
Unknown Object (File)
Feb 17 2024, 2:29 AM
Unknown Object (File)
Jan 14 2024, 5:58 AM
Unknown Object (File)
Jan 12 2024, 11:43 AM
Unknown Object (File)
Dec 20 2023, 6:46 AM
Unknown Object (File)
Nov 11 2023, 10:28 AM
Unknown Object (File)
Nov 8 2023, 12:55 AM
Unknown Object (File)
Nov 1 2023, 5:46 PM
Unknown Object (File)
Oct 10 2023, 9:27 AM

Details

Reviewers
manu
andrew
gonzo
Summary

This updates the documentation to match what will be supported for the upcoming 13 release.

Diff Detail

Repository
rD FreeBSD doc repository - subversion
Lint
Lint Skipped
Unit
Tests Skipped
Build Status
Buildable 31275

Event Timeline

jmg requested review of this revision.May 24 2020, 8:12 PM
jmg created this revision.

add the path to the diff...

arm.xml
90 ↗(On Diff #72197)

There is also A13, A31/A31S, A33, A83T and H3/H2PLUS

92 ↗(On Diff #72197)

Support was removed for those.

93 ↗(On Diff #72197)

BCM2836 is also supported (RPI2)

96 ↗(On Diff #72197)

Armada 500 isn't supported iirc, the only ARMv6 soc that we support is the broadcom one in the RPI1/0

99 ↗(On Diff #72197)

You can remove RK3188

100 ↗(On Diff #72197)

Support is broken and I have a review somewhere to remove it, I think you can drop this line.

en_US.ISO8859-1/htdocs/platforms/arm.xml
84

should we just drop this entire section? Made more sense in the past, but now, not so much. Also, I pulled the list from the ident CPUs, so we might not actually run on all of them.

90

We have conf for this, but not at all familiar if we actually run on this hardware.

151

did we even ever support the TS-7200? I can try to boot it, but we only have config file for 7800...

update with comments from manu

jmg marked 3 inline comments as not done.May 24 2020, 8:29 PM

these should not have been marked done. Why phab did this I have no clue. Maybe they think that if you update a patch you address all of your comments, but that is a TERRIBLE assumption to make.

jmg marked 6 inline comments as done.May 24 2020, 8:30 PM

a later diff has addressed these comments.

en_US.ISO8859-1/htdocs/platforms/arm.xml
73

s/Note/Not/

95

Not sure it's worth putting armv7 here, also both are so if we put it, put it at the end of the line.

val_packett.cool added inline comments.
en_US.ISO8859-1/htdocs/platforms/arm.xml
73

Note all? "Not all"?

84

Everything is forward compatible within each ARMvX generation, I agree that listing cores is silly.

90

I doubt that the hardware is branded "SOCFPGA", that sounds like an internal development term

91

note: Cavium has been acquired by Marvell, maybe say something like "Marvell/Cavium ThunderX..."

If we're going to list proper server-grade machines (SBSA) like the ThunderX2 here, there's a lot more.
I would really prefer separate sections for ACPI and embedded (FDT), something like this:

<h4>Generic ARMv8 Server/Workstation Support</h4>

FreeBSD supports <a href="https://developer.arm.com/architectures/platform-design/server-systems">Arm ServerReady</a> compliant systems that implement the Server Base System Architecture (SBSA) and the Server Base Boot Requirements (SBBR). These machines boot with UEFI and use ACPI for device configuration.
The following platforms have been confirmed to boot:

- Marvell/Cavium ThunderX (CN88xx), ThunderX2 (CN99xx)
- AMD Opteron A1100 ("Seattle")
- Ampere eMAG 8180
- Marvell Armada 8020/8040 (with EDK2 UEFI firmware)
- NXP Layerscape LX2160A (with EDK2 UEFI firmware)
- Amazon EC2 Graviton/Graviton2 virtual instances

<h4>Supported Embedded SoCs</h4>

- Rockchip RK3328/RK3399
- ...
95–96

"8k" is a very informal abbreviation.

"Marvell Armada 7020/7040/8020/8040", or just "Marvell Armada 8020/8040" if we're not that sure about 7k (they're the same family)

97

Would be nice to clarify or remove, I only saw like one driver for something Snapdragon-related.

129–130

really bad entry, probably about old *Samsung* Chromebooks

150

isn't "SheevaPlug" usually written without a space?

@jmg would you refresh this now after the Asciidoc conversion?