fixes to mul

Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>

azure-pipelines.yml

@@ -66,15 +66,6 @@ jobs:
       scriptPath: scripts/mk_unix_dist.py
       arguments: --nodotnet --nojava --arch=arm64
       pythonInterpreter: $(python)
-  - task: CopyFiles@2
-    inputs:
-      sourceFolder: dist
-      contents: '*.zip'
-      targetFolder: $(Build.ArtifactStagingDirectory)
-  - task: PublishPipelineArtifact@0
-    inputs:
-      artifactName: 'ManyLinuxBuildArm64'
-      targetPath: $(Build.ArtifactStagingDirectory)
 
 - job: "Ubuntu20OCaml"
   displayName: "Ubuntu 20 with OCaml"

src/ast/sls/bv_sls_eval.cpp

@@ -1002,10 +1002,11 @@ namespace bv {
     * 8*e = a*(2b), then a = 4e*b^-1
     */
     bool sls_eval::try_repair_mul(bvect const& e, bvval& a, bvect const& b) {
+        verbose_stream() << e << " := " << a << " * " << b << "\n";
         unsigned parity_e = a.parity(e);
         unsigned parity_b = a.parity(b);
 
-        if (a.is_zero(e)) {
+        if (a.is_zero(b)) {
             a.get_variant(m_tmp, m_rand);
             if (m_rand(10) != 0)
                 for (unsigned i = 0; i < b.bw - parity_b; ++i)
@@ -1018,6 +1019,8 @@ namespace bv {
             return a.set_repair(random_bool(), m_tmp);            
         }      
 
+
+
 #if 0
         verbose_stream() << "solve for " << e << "\n";