WIP

Signed-off-by: Hanno Becker <beckphan@amazon.co.uk>

Author: hanno-becker

dev/x86_64/src/poly_basemul.inc

@@ -50,7 +50,7 @@
  *    r = ac + bdz = r0 + r1,
  *    s = ad + bc = s0 + s1.
  */
-.macro schoolbook iter k
+.macro schoolbook iter, k
 vmovdqa (256*\k + 32*\iter +  0)*2(%rsi), _a
 vmovdqa (256*\k + 32*\iter + 16)*2(%rsi), _b
 vmovdqa (256*\k + 32*\iter +  0)*2(%rdx), _c
@@ -114,14 +114,14 @@ vmovdqa _s0, (32*\iter + 16)*2(%rdi)
 .endm
 
 .macro poly_basemul k
-schoolbook 0 \k
-schoolbook 1 \k
-schoolbook 2 \k
-schoolbook 3 \k
-schoolbook 4 \k
-schoolbook 5 \k
-schoolbook 6 \k
-schoolbook 7 \k
+schoolbook 0, \k
+schoolbook 1, \k
+schoolbook 2, \k
+schoolbook 3, \k
+schoolbook 4, \k
+schoolbook 5, \k
+schoolbook 6, \k
+schoolbook 7, \k
 .endm
 
 .macro polyvec_basemul k

mlkem/src/native/x86_64/src/poly_basemul.inc

@@ -50,7 +50,7 @@
  *    r = ac + bdz = r0 + r1,
  *    s = ad + bc = s0 + s1.
  */
-.macro schoolbook iter k
+.macro schoolbook iter, k
 vmovdqa (256*\k + 32*\iter +  0)*2(%rsi), _a
 vmovdqa (256*\k + 32*\iter + 16)*2(%rsi), _b
 vmovdqa (256*\k + 32*\iter +  0)*2(%rdx), _c
@@ -114,14 +114,14 @@ vmovdqa _s0, (32*\iter + 16)*2(%rdi)
 .endm
 
 .macro poly_basemul k
-schoolbook 0 \k
-schoolbook 1 \k
-schoolbook 2 \k
-schoolbook 3 \k
-schoolbook 4 \k
-schoolbook 5 \k
-schoolbook 6 \k
-schoolbook 7 \k
+schoolbook 0, \k
+schoolbook 1, \k
+schoolbook 2, \k
+schoolbook 3, \k
+schoolbook 4, \k
+schoolbook 5, \k
+schoolbook 6, \k
+schoolbook 7, \k
 .endm
 
 .macro polyvec_basemul k

scripts/autogen

@@ -2116,17 +2116,19 @@ def update_via_simpasm(
             # Add syntax option for x86_64
             if arch == "x86_64" and x86_64_syntax != "att":
                 cmd += ["--syntax", x86_64_syntax]
+            cmd += ["-v"]
             r = subprocess.run(
                 cmd,
-                stdout=subprocess.DEVNULL,
-                stderr=subprocess.PIPE,
+                stdout=subprocess.PIPE,
+                stderr=subprocess.STDOUT,
                 check=True,
                 text=True,
             )
+            print(r.stdout)
         except subprocess.CalledProcessError as e:
             print(f"Command failed: {' '.join(cmd)}")
             print(f"Exit code: {e.returncode}")
-            print(f"stderr: {e.stderr}")
+            print(f"stdout/stderr: {e.stdout}")
             raise Exception("Failed to run simpasm") from e
         tmp.seek(0)
         new_contents = tmp.read().decode()

scripts/simpasm

@@ -256,6 +256,8 @@ def simplify(logger, args, asm_input, asm_output=None):
         logger.debug(f"Disassembling temporary object file {tmp_objfile0} ...")
         disasm = run_cmd(cmd).stdout
 
+        print(disasm)
+
         logger.debug("Patching up disassembly ...")
         simplified = patchup_disasm(disasm, cfify=args.cfify)