arm64: dts: adi: sc846: fix DRAM size and drop SRAM memory nodes

[artursartamonovsadi]

PR Description

Set DDR memroy region to max memory.
Remove unused and untested L2 SRAM node
Set rpmsg node to match L2 memory region

Tested with:
stress --vm 2 --vm-bytes 64M --timeout 60s

PR Type

• Bug fix (a change that fixes an issue)
• New feature (a change that adds new functionality)
• Breaking change (a change that affects other repos or cause CIs to fail)

PR Checklist

• I have conducted a self-review of my own code changes
• I have compiled my changes, including the documentation
• I have tested the changes on the relevant hardware
• I have updated the documentation outside this repo accordingly
• I have provided links for the relevant upstream lore

[github-actions]

LLM review

This series fixes the SC846 SoM/EZ-KIT DRAM size and removes duplicated/unused memory nodes.

run: 32498102643

f5754cf8b - fix DRAM size and drop SRAM memory nodes

Verified: memory@80000000 is now correctly <0x80000000 0x40000000> (1GB),
the duplicated memory/reserved-memory nodes between sc846.dtsi and
sc846-som.dts are gone, and no other file references the removed labels
(sram1_res, sram1_mmap, rsc_tbl1, vdev0buffer, vdev1vrings,
vdev1buffer). Confirmed by compiling adi/sc846-som.dtb and
adi/sc846-som-ezkit.dtb and decompiling the result.

Possible inconsistency: the relocated vdev0vring0/rsc_tbl0 reserved
regions (now at 0x20400000/0x20404000) fall 2MB outside the unchanged
adi,l2-da = <0x20000000 0x20200000> window and the unchanged second reg
entry of the core1-rproc@28240000 node in sc846-som.dts. The rsc-table
lookup itself doesn't require being inside adi,l2-da (it's resolved via
of_reserved_mem_lookup()/devm_ioremap_wc(), see
drivers/remoteproc/adi_remoteproc.c), but since the commit message states
these are now the "actual memory ranges", it would help to also confirm/update
adi,l2-da and reg[1] so the DT is self-consistent about the SHARC core's
usable memory window. No public datasheet for SC846 is available yet (checked
the ADI docling PDF sitemap), so this could only be checked for internal
consistency, not against a datasheet - flagging for maintainer/hardware
confirmation rather than guessing a fix.

Also note (pre-existing, not introduced by this commit): the core1-rproc
node still has no vdev-vring/memory-region property, so the vdev0vrings
label is an orphan phandle either way.

CI warnings

The only dtc warning reproduced (unique_unit_address on
arch/arm64/boot/dts/adi/sc846.dtsi:218, syscon@31004600 vs
pinctrl0@31004600) is pre-existing and already documented as deliberate in
the dtsi; unrelated to this commit.

Verification data

No SC846/SC59x datasheet is currently published in the ADI docling mirror
(en-pdf-sitemap.xml), so the reserved-memory relocation could not be
checked against a hardware reference manual. Verification was instead done
by: building the dtb (set_arch gcc_aarch64, make defconfig, make adi/sc846-som.dtb adi/sc846-som-ezkit.dtb), decompiling it with
scripts/dtc/dtc, running checkpatch.pl, and running dt-validate
(dtschema) against the compiled dtb, then cross-referencing the
adi,l2-da/adi,rsc-table consumers in drivers/remoteproc/adi_remoteproc.c.

Suggested patches

No fixup patches are included this round: the one open question (l2-da/reg
window vs. new rsc-table address) needs hardware/firmware confirmation before
a correct value can be proposed with confidence.