[PATCH v3] msvcrt: Compile but show an error if Bessel functions aren't available

[Alex Henrie]

Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com>
---
Fixes https://bugs.winehq.org/show_bug.cgi?id=37809

v3: Dropped incorrect fallbacks for j0 and j1

 configure.ac       |  8 +++++++-
 dlls/msvcrt/math.c | 30 ++++++++++++++++++++++++++++++
 2 files changed, 37 insertions(+), 1 deletion(-)

diff --git a/configure.ac b/configure.ac
index 2e99320371..e198018427 100644
--- a/configure.ac
+++ b/configure.ac
@@ -2698,6 +2698,9 @@ AC_CHECK_FUNCS(\
 	exp2f \
 	expm1 \
 	expm1f \
+	j0 \
+	j1 \
+	jn \
 	lgamma \
 	lgammaf \
 	llrint \
@@ -2722,7 +2725,10 @@ AC_CHECK_FUNCS(\
 	round \
 	roundf \
 	trunc \
-	truncf
+	truncf \
+	y0 \
+	y1 \
+	yn
 )
 LIBS="$ac_save_LIBS"
 
diff --git a/dlls/msvcrt/math.c b/dlls/msvcrt/math.c
index 9b480117fc..cc1979b96c 100644
--- a/dlls/msvcrt/math.c
+++ b/dlls/msvcrt/math.c
@@ -1412,7 +1412,12 @@ INT CDECL MSVCRT__isnan(double num)
 double CDECL MSVCRT__j0(double num)
 {
   /* FIXME: errno handling */
+#ifdef HAVE_J0
   return j0(num);
+#else
+  FIXME( "not implemented\n" );
+  return 0;
+#endif
 }
 
 /*********************************************************************
@@ -1421,7 +1426,12 @@ double CDECL MSVCRT__j0(double num)
 double CDECL MSVCRT__j1(double num)
 {
   /* FIXME: errno handling */
+#ifdef HAVE_J1
   return j1(num);
+#else
+  FIXME( "not implemented\n" );
+  return 0;
+#endif
 }
 
 /*********************************************************************
@@ -1430,7 +1440,12 @@ double CDECL MSVCRT__j1(double num)
 double CDECL MSVCRT__jn(int n, double num)
 {
   /* FIXME: errno handling */
+#ifdef HAVE_JN
   return jn(n, num);
+#else
+  FIXME( "not implemented\n" );
+  return 0;
+#endif
 }
 
 /*********************************************************************
@@ -1440,12 +1455,17 @@ double CDECL MSVCRT__y0(double num)
 {
   double retval;
   if (!isfinite(num)) *MSVCRT__errno() = MSVCRT_EDOM;
+#ifdef HAVE_Y0
   retval  = y0(num);
   if (MSVCRT__fpclass(retval) == MSVCRT__FPCLASS_NINF)
   {
     *MSVCRT__errno() = MSVCRT_EDOM;
     retval = sqrt(-1);
   }
+#else
+  FIXME( "not implemented\n" );
+  retval = 0;
+#endif
   return retval;
 }
 
@@ -1456,12 +1476,17 @@ double CDECL MSVCRT__y1(double num)
 {
   double retval;
   if (!isfinite(num)) *MSVCRT__errno() = MSVCRT_EDOM;
+#ifdef HAVE_Y1
   retval  = y1(num);
   if (MSVCRT__fpclass(retval) == MSVCRT__FPCLASS_NINF)
   {
     *MSVCRT__errno() = MSVCRT_EDOM;
     retval = sqrt(-1);
   }
+#else
+  FIXME( "not implemented\n" );
+  retval = 0;
+#endif
   return retval;
 }
 
@@ -1472,12 +1497,17 @@ double CDECL MSVCRT__yn(int order, double num)
 {
   double retval;
   if (!isfinite(num)) *MSVCRT__errno() = MSVCRT_EDOM;
+#ifdef HAVE_YN
   retval  = yn(order,num);
   if (MSVCRT__fpclass(retval) == MSVCRT__FPCLASS_NINF)
   {
     *MSVCRT__errno() = MSVCRT_EDOM;
     retval = sqrt(-1);
   }
+#else
+  FIXME( "not implemented\n" );
+  retval = 0;
+#endif
   return retval;
 }
 
-- 
2.16.2

[Piotr Caban]

Hi Alex,

On 03/02/18 17:21, Alex Henrie wrote:
> @@ -1472,12 +1497,17 @@ double CDECL MSVCRT__yn(int order, double num)
>   {
>     double retval;
>     if (!isfinite(num)) *MSVCRT__errno() = MSVCRT_EDOM;
> +#ifdef HAVE_YN
>     retval  = yn(order,num);
>     if (MSVCRT__fpclass(retval) == MSVCRT__FPCLASS_NINF)
>     {
>       *MSVCRT__errno() = MSVCRT_EDOM;
>       retval = sqrt(-1);
>     }
> +#else
> +  FIXME( "not implemented\n" );
> +  retval = 0;
> +#endif
>     return retval;
>   }
> The patch looks mostly good for me. Could you please also remove the 
> spaces after '(' and before ')' in FIXME to match surrounding code 
> style? While you're at it, could you please also remove sqrt(-1) calls 
> on error paths and just return NAN?

Thanks,
Piotr