[gcc r12-7184] analyzer: handle more casts of string literals [PR98797]

[David Malcolm <dmalcolm@gcc.gnu.org>]

https://gcc.gnu.org/g:2ac7b19f1e9219f46ccf55f25d8acb3e02e9a2d4

commit r12-7184-g2ac7b19f1e9219f46ccf55f25d8acb3e02e9a2d4
Author: David Malcolm <dmalcolm@redhat.com>
Date:   Wed Feb 9 19:06:15 2022 -0500

    analyzer: handle more casts of string literals [PR98797]
    
    gcc/analyzer/ChangeLog:
            PR analyzer/98797
            * region-model-manager.cc
            (region_model_manager::maybe_fold_sub_svalue): Generalize getting
            individual chars of a STRING_CST from element_region to any
            subregion which is a concrete access of a single byte from its
            parent region.
            * region.cc (region::get_relative_concrete_byte_range): New.
            * region.h (region::get_relative_concrete_byte_range): New decl.
    
    gcc/testsuite/ChangeLog:
            PR analyzer/98797
            * gcc.dg/analyzer/casts-1.c: Mark xfails as fixed; add further
            test coverage for casts of string literals.
    
    Signed-off-by: David Malcolm <dmalcolm@redhat.com>

Diff:
---
 gcc/analyzer/region-model-manager.cc    | 19 +++++++++-----
 gcc/analyzer/region.cc                  | 28 ++++++++++++++++++++
 gcc/analyzer/region.h                   |  6 +++++
 gcc/testsuite/gcc.dg/analyzer/casts-1.c | 45 +++++++++++++++++++++++++++------
 4 files changed, 84 insertions(+), 14 deletions(-)

diff --git a/gcc/analyzer/region-model-manager.cc b/gcc/analyzer/region-model-manager.cc
index 010ad078849..d7156c5499f 100644
--- a/gcc/analyzer/region-model-manager.cc
+++ b/gcc/analyzer/region-model-manager.cc
@@ -782,15 +782,22 @@ region_model_manager::maybe_fold_sub_svalue (tree type,
   /* Handle getting individual chars from a STRING_CST.  */
   if (tree cst = parent_svalue->maybe_get_constant ())
     if (TREE_CODE (cst) == STRING_CST)
-      if (const element_region *element_reg
-	    = subregion->dyn_cast_element_region ())
-	{
-	  const svalue *idx_sval = element_reg->get_index ();
-	  if (tree cst_idx = idx_sval->maybe_get_constant ())
+      {
+	/* If we have a concrete 1-byte access within the parent region... */
+	byte_range subregion_bytes (0, 0);
+	if (subregion->get_relative_concrete_byte_range (&subregion_bytes)
+	    && subregion_bytes.m_size_in_bytes == 1)
+	  {
+	    /* ...then attempt to get that char from the STRING_CST.  */
+	    HOST_WIDE_INT hwi_start_byte
+	      = subregion_bytes.m_start_byte_offset.to_shwi ();
+	    tree cst_idx
+	      = build_int_cst_type (size_type_node, hwi_start_byte);
 	    if (const svalue *char_sval
 		= maybe_get_char_from_string_cst (cst, cst_idx))
 	      return get_or_create_cast (type, char_sval);
-	}
+	  }
+      }
 
   if (const initial_svalue *init_sval
 	= parent_svalue->dyn_cast_initial_svalue ())
diff --git a/gcc/analyzer/region.cc b/gcc/analyzer/region.cc
index 0adc75e577d..5ac24fb9f9b 100644
--- a/gcc/analyzer/region.cc
+++ b/gcc/analyzer/region.cc
@@ -539,6 +539,34 @@ region::get_relative_concrete_offset (bit_offset_t *) const
   return false;
 }
 
+/* Attempt to get the position and size of this region expressed as a
+   concrete range of bytes relative to its parent.
+   If successful, return true and write to *OUT.
+   Otherwise return false.  */
+
+bool
+region::get_relative_concrete_byte_range (byte_range *out) const
+{
+  /* We must have a concrete offset relative to the parent.  */
+  bit_offset_t rel_bit_offset;
+  if (!get_relative_concrete_offset (&rel_bit_offset))
+    return false;
+  /* ...which must be a whole number of bytes.  */
+  if (rel_bit_offset % BITS_PER_UNIT != 0)
+    return false;
+  byte_offset_t start_byte_offset = rel_bit_offset / BITS_PER_UNIT;
+
+  /* We must have a concrete size, which must be a whole number
+     of bytes.  */
+  byte_size_t num_bytes;
+  if (!get_byte_size (&num_bytes))
+    return false;
+
+  /* Success.  */
+  *out = byte_range (start_byte_offset, num_bytes);
+  return true;
+}
+
 /* Dump a description of this region to stderr.  */
 
 DEBUG_FUNCTION void
diff --git a/gcc/analyzer/region.h b/gcc/analyzer/region.h
index 53112175266..2f987e49fa8 100644
--- a/gcc/analyzer/region.h
+++ b/gcc/analyzer/region.h
@@ -182,6 +182,12 @@ public:
      Otherwise return false.  */
   virtual bool get_relative_concrete_offset (bit_offset_t *out) const;
 
+  /* Attempt to get the position and size of this region expressed as a
+     concrete range of bytes relative to its parent.
+     If successful, return true and write to *OUT.
+     Otherwise return false.  */
+  bool get_relative_concrete_byte_range (byte_range *out) const;
+
   void
   get_subregions_for_binding (region_model_manager *mgr,
 			      bit_offset_t start_bit_offset,
diff --git a/gcc/testsuite/gcc.dg/analyzer/casts-1.c b/gcc/testsuite/gcc.dg/analyzer/casts-1.c
index 15cd85f77cf..7e4af384971 100644
--- a/gcc/testsuite/gcc.dg/analyzer/casts-1.c
+++ b/gcc/testsuite/gcc.dg/analyzer/casts-1.c
@@ -13,6 +13,14 @@ struct s2
   char arr[4];
 };
 
+struct s3
+{
+  struct inner {
+    char a;
+    char b;
+  } arr[2];
+};
+
 void test_1 ()
 {
   struct s1 x = {'A', 'B', 'C', 'D'};
@@ -24,10 +32,16 @@ void test_1 ()
   __analyzer_eval (((struct s2 *)&x)->arr[1] == 'B'); /* { dg-warning "TRUE" } */
   __analyzer_eval (((struct s2 *)&x)->arr[2] == 'C'); /* { dg-warning "TRUE" } */
   __analyzer_eval (((struct s2 *)&x)->arr[3] == 'D'); /* { dg-warning "TRUE" } */
+  struct s3 *p3 = (struct s3 *)&x;
+  __analyzer_eval (p3->arr[0].a == 'A'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p3->arr[0].b == 'B'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p3->arr[1].a == 'C'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p3->arr[1].b == 'D'); /* { dg-warning "TRUE" } */
 
   ((struct s2 *)&x)->arr[1] = '#';
   __analyzer_eval (((struct s2 *)&x)->arr[1] == '#'); /* { dg-warning "TRUE" } */
   __analyzer_eval (x.b == '#'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p3->arr[0].b == '#'); /* { dg-warning "TRUE" } */
 }
 
 void test_2 ()
@@ -38,12 +52,27 @@ void test_2 ()
   __analyzer_eval (x.arr[2] == 'C'); /* { dg-warning "TRUE" } */
   __analyzer_eval (x.arr[3] == 'D'); /* { dg-warning "TRUE" } */
   struct s1 *p = (struct s1 *)&x;
-  __analyzer_eval (p->a == 'A'); /* { dg-warning "TRUE" "true" { xfail *-*-* } } */
-  /* { dg-bogus "UNKNOWN" "unknown" { xfail *-*-* } .-1 } */
-  __analyzer_eval (p->b == 'B'); /* { dg-warning "TRUE" "true" { xfail *-*-* } } */
-  /* { dg-bogus "UNKNOWN" "unknown" { xfail *-*-* } .-1 } */
-  __analyzer_eval (p->c == 'C'); /* { dg-warning "TRUE" "true" { xfail *-*-* } } */
-  /* { dg-bogus "UNKNOWN" "unknown" { xfail *-*-* } .-1 } */
-  __analyzer_eval (p->d == 'D'); /* { dg-warning "TRUE" "true" { xfail *-*-* } } */
-  /* { dg-bogus "UNKNOWN" "unknown" { xfail *-*-* } .-1 } */
+  __analyzer_eval (p->a == 'A'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p->b == 'B'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p->c == 'C'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p->d == 'D'); /* { dg-warning "TRUE" } */
+}
+
+void test_3 ()
+{
+  struct s3 x = {'A', 'B', 'C', 'D'};
+  __analyzer_eval (x.arr[0].a == 'A'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (x.arr[0].b == 'B'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (x.arr[1].a == 'C'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (x.arr[1].b == 'D'); /* { dg-warning "TRUE" } */
+  struct s1 *p1 = (struct s1 *)&x;
+  __analyzer_eval (p1->a == 'A'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p1->b == 'B'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p1->c == 'C'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p1->d == 'D'); /* { dg-warning "TRUE" } */
+  struct s2 *p2 = (struct s2 *)&x;
+  __analyzer_eval (p2->arr[0] == 'A'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p2->arr[1] == 'B'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p2->arr[2] == 'C'); /* { dg-warning "TRUE" } */
+  __analyzer_eval (p2->arr[3] == 'D'); /* { dg-warning "TRUE" } */
 }