feat(linux): Add Ethernet performance numbers for 12.01 release

[danish-ti]

Add netperf Ethernet performance data for AM62X, AM62PX, AM62LX and AM64X in their respective performance guide RST files. Each section covers CPSW TCP bidirectional throughput (with and without interrupt pacing) and UDP egress/ingress 0-loss throughput across 64/128/256/ 1024/1518B frame sizes. AM64X additionally includes ICSSG results.

[github-actions]

New warnings found with rstcheck:

source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:507: (ERROR/3) Inconsistent title style: skip from level 4 to 6.
source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:521: (ERROR/3) Inconsistent title style: skip from level 4 to 6.
source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:535: (ERROR/3) Inconsistent title style: skip from level 4 to 6.

[StaticRocket]

Hm. Don't like this sudden formatting change. What's proposing it?

[danish-ti]

This comes from a heading-level shift introduced by cc2de1e feat(linux): AM62X: Add initial performance guide for 12.01.00 release PR #758 , which rewrote the file and dropped the = heading level. In 12.00 performance doc, the Ethernet section used = (level 3) → - (level 4, CPSW) → ^ (level 5, TCP/UDP sub-headings). After cc2de1e, - became level 3 and ^ became level 4. When restoring the Ethernet section, CPSW/ICSSG now sits at ^ (level 4), so the TCP/UDP sub-headings need a new level 5 — hence ~. The character itself is arbitrary; the level shift was caused by the 12.01 template rewrite, not this PR. Level 5 can be either " or ~.

[StaticRocket]

Oh. This file has sections all out of order.

[StaticRocket]

Ugh, alright. I'll look at fixing those good awful templates again. Change the am64 document to match the others for now, assuming that makes the TOC tree happy. Test it locally and make sure it's not trying to put that section above anything else in the hierarchy.

[danish-ti]

Ugh, alright. I'll look at fixing those good awful templates again. Change the am64 document to match the others for now, assuming that makes the TOC tree happy. Test it locally and make sure it's not trying to put that section above anything else in the hierarchy.

@StaticRocket If I change the AM64 document to match the others, rstcheck fails with "Title level inconsistent" erros.
source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:507: (SEVERE/4) Title level inconsistent: source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:521: (SEVERE/4) Title level inconsistent: source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:535: (SEVERE/4) Title level inconsistent:
This happens because AM64 already uses " for Level 5 headings. Changing AM64 breaks its existing hierarchy. To fix this and keep everything consistent across the PR, I suggest we update the AM62 files (am62x, am62l, am62p) to use " for Level 5 instead. This resolves the warnings and standardises the templates.

[StaticRocket]

The problem with that is, that requires more changes that go explicitly against the section hierarchy outlined in the contribution guide.

[danish-ti]

@StaticRocket

1. Keeping " as level 5 for both AM64 and AM62 docs requires more changes that go explicitly against the section hierarchy outlined in the contribution guide.
2. Keeping ~ as level 5 for both AM64 and AM62 docs introduces rstcheck warnings. source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:507: (SEVERE/4) Title level inconsistent: source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:521: (SEVERE/4) Title level inconsistent: source/devices/AM64X/linux/RT_Linux_Performance_Guide.rst:535: (SEVERE/4) Title level inconsistent:
3. Keeping ~ in AM62x and " in AM64x satisfies both 1 and 2 but creates inconsistency across docs.
4. Implement point 2 but to avoid rstwarnings, we need to change EMMC EXT4 and other level 5 from " to ~. I didn't want to touch anything not related to networking in this PR.

What do you think should be done here then? The current state of PR is 3.

[StaticRocket]

Eh. If the linter is happy for now then it's fine as is. I have a pending PR reworking that template file again.