Skip to content

Ghidra is a software reverse engineering (SRE) framework

hexagon: refresh comments to reflect current support

Shawn Hoffman0345ff83cfd8
diff --git a/Ghidra/Processors/Hexagon/data/languages/hexagon.sinc b/Ghidra/Processors/Hexagon/data/languages/hexagon.sinc
index 3437c42c8c..00d90c7820 100755
--- a/Ghidra/Processors/Hexagon/data/languages/hexagon.sinc
+++ b/Ghidra/Processors/Hexagon/data/languages/hexagon.sinc
@@ -1,4 +1,4 @@
-# Qualcomm Hexagon (V73) General Instruction Set
+# Qualcomm Hexagon (V60-V81) General Instruction Set
#
# Custom pcode-op
diff --git a/Ghidra/Processors/Hexagon/data/languages/hexagon.slaspec b/Ghidra/Processors/Hexagon/data/languages/hexagon.slaspec
index fa67708196..1d1e25f58e 100755
--- a/Ghidra/Processors/Hexagon/data/languages/hexagon.slaspec
+++ b/Ghidra/Processors/Hexagon/data/languages/hexagon.slaspec
@@ -1,12 +1,14 @@
# Qualcomm Hexagon (V79) and HVX (V79)
## KNOWN ISSUES
-# 1. Shift amounts may be positive or negative, however in some cases no special handling
-# is provided for negative shift amounts.
+# 1. Shift amounts may be positive or negative. The register-form shifts model the
+# 7-bit signed shift amount explicitly; other forms may still assume a positive amount.
# 2. There are many complex instructions with unimplemented pcode or simple custom pcodeops
#
## NOTES
-# 1. Implementation includes V73 system registers, instruction set may be incomplete
+# 1. Implementation includes V73 system registers, instruction set may be incomplete.
+# Decode additionally covers the V66 ZReg matrix extension and V81 valign4 as
+# decode-only stubs (unimpl pcode).
# 2. HVX register pairs may be access with a vector-swap mode (lsb of 5-bit field is 1)
# introduced with Hexagon v67.
# 3. HVX vector register size is 128-bytes (see defines below), paired size is 256-bytes,
diff --git a/Ghidra/Processors/Hexagon/data/languages/hexagon_hvx.sinc b/Ghidra/Processors/Hexagon/data/languages/hexagon_hvx.sinc
index 43509bb90b..35a3daf4a8 100644
--- a/Ghidra/Processors/Hexagon/data/languages/hexagon_hvx.sinc
+++ b/Ghidra/Processors/Hexagon/data/languages/hexagon_hvx.sinc
@@ -2,6 +2,8 @@
# NOTES:
# - HVX implementation has been updated through V79
+# - Decode additionally covers the V66 ZReg matrix family (vrmpyz / vr8mpyz / vr16mpyz
+# and the z-register loads) plus V81 valign4, as decode-only stubs (unimpl pcode).
# - HVX hardware supports either 64-byte or 128-byte vector lengths. See hexagon.slaspec for
# default length of 128-bytes.