================
@@ -19,31 +19,29 @@
# TODO: more exhaustive testing of immediate encoding.
-# CHECK-ASM-AND-OBJ: c.ldsp ra, 0(sp)
-# CHECK-ASM: encoding: [0x82,0x60]
+# CHECK-ASM-AND-OBJ: c.ldsp s0, 0(sp)
+# CHECK-ASM: encoding: [0x02,0x64]
# CHECK-NO-EXT: error: instruction requires the following: 'C' (Compressed
Instructions) or 'Zca' (part of the C extension, excluding compressed floating
point loads/stores){{$}}
-# CHECK-NO-RV64: error: instruction requires the following: RV64I Base
Instruction Set{{$}}
-c.ldsp ra, 0(sp)
-# CHECK-ASM-AND-OBJ: c.sdsp ra, 504(sp)
-# CHECK-ASM: encoding: [0x86,0xff]
+# CHECK-NO-RV64: error: instruction requires the following: 'Zclsd'
(Compressed Load/Store pair instructions){{$}}
----------------
topperc wrote:
I don't think there's any way to fix this. The system is going to print the
missing features from the closest matching instruction. It isn't aggregating
across possible matches.
I think its uncommon for people to mistakenly compiling code using RV32 when
they meant RV64 or vice versa. We can probably trust that their used the right
base.
https://github.com/llvm/llvm-project/pull/131094
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
