[clang] Fix handling of "%zd" format specifier in scanf

[alexander-shaposhnikov]

Summary

This is a follow-up for https://reviews.llvm.org/D35427
ScanfFormatString.cpp has the same issue (incorrect handling of
the specifiers for ssize_t) and this diff fixes it and adds tests.

Test plan: make check-all

[alexander-shaposhnikov]

test/Sema/format-strings-fixit-ssize_t.c

4

c99 is necessary for "zd"

test/Sema/format-strings-scanf.c

1

c11 is necessary for _Generic below.
I have not found a more reliable way to define ssize_t
which doesn't require a large number of #ifdef ...

[aaron.ballman]

LGTM with a small testing nit.

test/Sema/format-strings-scanf.c

1

Ugh, that is pretty nasty, but I cannot think of a better way to define it in a generic manner.

198

Can you add a passing test for %zn?

[alexander-shaposhnikov]

Address comments

lib/Analysis/ScanfFormatString.cpp

case ConversionSpecifier::iArg:
         case LengthModifier::AsLongLong:
         case LengthModifier::AsQuad:
           return ArgType::PtrTo(Ctx.LongLongTy);
         case LengthModifier::AsInt64:
           return ArgType::PtrTo(ArgType(Ctx.LongLongTy, "__int64"));
         case LengthModifier::AsIntMax:
           return ArgType::PtrTo(ArgType(Ctx.getIntMaxType(), "intmax_t"));
         case LengthModifier::AsSizeT:
-          // FIXME: ssize_t.
-          return ArgType();
+          return ArgType::PtrTo(ArgType(Ctx.getSignedSizeType(), "ssize_t"));
         case LengthModifier::AsPtrDiff:
           return ArgType::PtrTo(ArgType(Ctx.getPointerDiffType(), "ptrdiff_t"));
         case LengthModifier::AsLongDouble:
           // GNU extension.
           return ArgType::PtrTo(Ctx.LongLongTy);
         case LengthModifier::AsAllocate:
         case LengthModifier::AsMAllocate:
         case LengthModifier::AsInt32:
case ConversionSpecifier::nArg:
         case LengthModifier::AsLongLong:
         case LengthModifier::AsQuad:
           return ArgType::PtrTo(Ctx.LongLongTy);
         case LengthModifier::AsInt64:
           return ArgType::PtrTo(ArgType(Ctx.LongLongTy, "__int64"));
         case LengthModifier::AsIntMax:
           return ArgType::PtrTo(ArgType(Ctx.getIntMaxType(), "intmax_t"));
         case LengthModifier::AsSizeT:
-          return ArgType(); // FIXME: ssize_t
+          return ArgType::PtrTo(ArgType(Ctx.getSignedSizeType(), "ssize_t"));
         case LengthModifier::AsPtrDiff:
           return ArgType::PtrTo(ArgType(Ctx.getPointerDiffType(), "ptrdiff_t"));
         case LengthModifier::AsLongDouble:
           return ArgType(); // FIXME: Is this a known extension?
         case LengthModifier::AsAllocate:
         case LengthModifier::AsMAllocate:
         case LengthModifier::AsInt32:
         case LengthModifier::AsInt3264:

test/Sema/format-strings-fixit-ssize_t.c

 // RUN: cp %s %t
-// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -pedantic -Wall -fixit %t
-// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -fsyntax-only -pedantic -Wall -Werror %t
-// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -E -o - %t | FileCheck %s
+// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -std=c99 -pedantic -Wall -fixit %t
+// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -std=c99 -fsyntax-only -pedantic -Wall -Werror %t
+// RUN: %clang_cc1 -triple x86_64-apple-darwin10 -std=c99 -E -o - %t | FileCheck %s
 
 /* This is a test of the various code modification hints that are
    provided as part of warning or extension diagnostics. All of the
    warnings will be fixed by -fixit, and the resulting file should
    compile cleanly with -Werror -pedantic. */
 
 int printf(char const *, ...);
+int scanf(const char *, ...);
 
 void test() {
   typedef signed long int ssize_t;
   printf("%f", (ssize_t) 42);
+  ssize_t s;
+  scanf("%f",  &s);
 }
 
 // CHECK: printf("%zd", (ssize_t) 42);
+// CHECK: scanf("%zd", &s)

test/Sema/format-strings-scanf.c

-// RUN: %clang_cc1 -fsyntax-only -verify -Wformat-nonliteral %s
+// RUN: %clang_cc1 -std=c11 -fsyntax-only -verify -Wformat-nonliteral %s
 
 // Test that -Wformat=0 works:
-// RUN: %clang_cc1 -fsyntax-only -Werror -Wformat=0 %s
+// RUN: %clang_cc1 -std=c11 -fsyntax-only -Werror -Wformat=0 %s
 
 #include <stdarg.h>
-typedef __typeof(sizeof(int)) size_t;
+typedef __SIZE_TYPE__ size_t;
+#define __SSIZE_TYPE__                                                         \
+  __typeof__(_Generic((__SIZE_TYPE__)0,                                        \
+                      unsigned long long int : (long long int)0,               \
+                      unsigned long int : (long int)0,                         \
+                      unsigned int : (int)0,                                   \
+                      unsigned short : (short)0,                               \
+                      unsigned char : (signed char)0))
+typedef __SSIZE_TYPE__ ssize_t;                         
 typedef struct _FILE FILE;
 typedef __WCHAR_TYPE__ wchar_t;
 
 int fscanf(FILE * restrict, const char * restrict, ...) ;
 int scanf(const char * restrict, ...) ;
 int sscanf(const char * restrict, const char * restrict, ...) ;
 int my_scanf(const char * restrict, ...) __attribute__((__format__(__scanf__, 1, 2)));
 
void test_qualifiers(const int *cip, volatile int* vip,
   scanf("%c", vcp); // No warning.
 
   typedef int* ip_t;
   typedef const int* cip_t;
   scanf("%d", (ip_t)0); // No warning.
   scanf("%d", (cip_t)0); // expected-warning{{format specifies type 'int *' but the argument has type 'cip_t' (aka 'const int *')}}
 }
 
+void test_size_types() {
+  size_t s = 0;
+  scanf("%zu", &s); // No warning.
+
+  double d1 = 0.;
+  scanf("%zu", &d1); // expected-warning-re{{format specifies type 'size_t *' (aka '{{.+}}') but the argument has type 'double *'}}
+
+  ssize_t ss = 0;
+  scanf("%zd", &s); // No warning.
+
+  double d2 = 0.;
+  scanf("%zd", &d2); // expected-warning-re{{format specifies type 'ssize_t *' (aka '{{.+}}') but the argument has type 'double *'}}
+
+  double d3 = 0.;
+  scanf("%zn", &d3); // expected-warning-re{{format specifies type 'ssize_t *' (aka '{{.+}}') but the argument has type 'double *'}}
+}
+
 void check_conditional_literal(char *s, int *i) {
   scanf(0 ? "%s" : "%d", i); // no warning
   scanf(1 ? "%s" : "%d", i); // expected-warning{{format specifies type 'char *'}}
   scanf(0 ? "%d %d" : "%d", i); // no warning
   scanf(1 ? "%d %d" : "%d", i); // expected-warning{{more '%' conversions than data arguments}}
   scanf(0 ? "%d %d" : "%d", i, s); // expected-warning{{data argument not used}}
   scanf(1 ? "%d %s" : "%d", i, s); // no warning
   scanf(i ? "%d %s" : "%d", i, s); // no warning
   scanf(i ? "%d" : "%d", i, s); // expected-warning{{data argument not used}}
   scanf(i ? "%s" : "%d", s); // expected-warning{{format specifies type 'int *'}}
 }