catspeed-cc/openwrt
1
1
Fork 0
Code

brain +15 wrinkles from inline thinking breakthrough - re: reverse engineering / parallel reconstruction

Browse Source
This commit is contained in:
mooleshacat 2026-06-15 20:31:30 -04:00
parent ea18b36d36
commit fae78f3823
Signed by: mooleshacat
GPG Key ID: 6F42FE1A481818C2
1 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
target/linux/ipq40xx/files-6.12/arch/arm/boot/dts/qcom/qcom-ipq4019-tew-829dru.dts
View File

@ -528,7 +528,7 @@ edma {
Then OpenWRT handles the communication between wan1, lan, wan2 based on it's firewall rules, etc. with physical separation between the security concerns as expected for an SMB device.
### ADDRESS SPACE CONFLICTS ###
NOTE: hardware addresses *MUST* match the mfr, other config - eg. daisy chaining - is optional, provided both switches support them (qcom-ipq4019.dtsi shows that both do)
NOTE: hardware addresses *MUST* match the mfr, other config - eg. daisy chaining - is optional, provided both switches support them (qcom-ipq4019.dtsi shows that both do support daisy chaining)
qcom-ipq4019.dtsi (upstream) stock-fixed.dts (mfr)
gmac: ethernet@c080000 -> edma@c080000 Delete upstream nodes, redefine as per mfr.
@ -558,9 +558,12 @@ edma {
If qcom-ipq4019.dtsi defines both a QCA8337 & QCA8075 presumably they would work in the daisy chained configuration - they are the same switches after all ...
Perhaps we go the daisy chain route, even though mfr does not? What are effects of this? It would still give us 8 LAN ports. The hardware in both cases - upstream and mfr - is the same (QCA8337, QCA8075)
YES, qcom-ipq4019.dtsi defines *BOTH* QCA8337 & QCA8075! That means we possibly can use their implementation instead, the switches are the exact same!
but the hardware address might be different ... upstream: switch@c000000 (QCA8337) & mdio@90000 (QCA8075)
YES, qcom-ipq4019.dtsi defines *BOTH* QCA8337 & QCA8075 in daisy chain config! That means we possibly can use their implementation instead, the switches are the exact same!
but the hardware address might be different ... upstream: switch@c000000 (QCA8337) / mfr: switch@???? (QCA8337)
we can use them but must make sure hardware lies at these exact addresses.
Note to self: Now we are getting into an interesting area of reverse engineering: parallel reconstruction.
We are utilizing the known possible configuration for the exact same hardware from upstream to make conclusions regarding the mfr configuration
ANSWER: mdio@90000 in upstream matches mdio@90000 in mfr and is a switch. This must be the QCA8075
that leaves the unknown switch to be the QCA8337