RE: [RFA] testsuite: Add a test for passing of environment variables to inferior

["Pierre Muller" <pierre.muller@ics-cnrs.unistra.fr>]

> -----Message d'origine-----
> De : gdb-patches-owner@sourceware.org [mailto:gdb-patches-
> owner@sourceware.org] De la part de Corinna Vinschen
> Envoyé : mardi 4 octobre 2011 15:45
> À : gdb-patches@sourceware.org
> Objet : Re: [RFA] testsuite: Add a test for passing of environment variables
> to inferior
> 
> On Oct  4 14:36, Pierre Muller wrote:
> >   Following Eli's comment that it was not clear
> > what I wanted to achieve, I wrote a new test.
> >
> >   Manually checking CVS GDB for mingw, Eli's patch
> > gives the correct output.
> >   But Stock Cygwin GDB currently passes none of these
> > TEST_GDB_XXX variables to inferior.
> >   Even the patch that I sent earlier is not correct:
> > it still fails for the last test,
> > once TEST_GDB_VAR1 has been set into GDB environment list,
> > it doesn't get removed on the last start of the inferior...
> >
> >   Corinna, I think this is the reason why I wanted to restore
> > the original environment layout (to avoid leaving unset
> > variables around.)
> 
> Yes, that would be necessary.  I'm wondering if we can't just utilze the
> global environ variable for this and spare us all the hassle.  Something
> along these lines:
> 
>   char **old_env = environ;
>   environ = in_env;
>   cygwin_internal (CW_SYNC_WINENV);
>   CreateProcessW (NULL environment pointer);
>   environ = old_env;

  I checked it out, but it still
fails for the last test, the missing TEST_GDB_VAR1
variable doesn't get removed from GDB environment list
apparently (This is probably related to the internals
of cygwin_internal (CW_SYNC_WINENV), no?

  Manually resetting all variables should help...
The patch below does not generate any failure on my test,
but I am still not sure what happens if you try to
remove an environment variable that was set when GDB started...
  Following Elis's comment, I also checked what happens
if I only call cygwin_internal (CW_SYNC_WINENV)
after resetting environ to the GDB value, but
the TEST_GDB_VAR1 doesn't get removed and the last test
fails.

Pierre
ChangeLog entry:
2011-10-04  Pierre Muller  <muller@ics.u-strasbg.fr>

	* windows-nat.c (windows_create_inferior): Handle in_env
	array for Cygwin GDB.

Index: windows-nat.c
===================================================================
RCS file: /cvs/src/src/gdb/windows-nat.c,v
retrieving revision 1.219
diff -u -p -r1.219 windows-nat.c
--- windows-nat.c       28 Sep 2011 09:07:54 -0000      1.219
+++ windows-nat.c       4 Oct 2011 14:40:59 -0000
@@ -1980,8 +1980,9 @@ windows_create_inferior (struct target_o
   cygwin_buf_t *toexec;
   cygwin_buf_t *cygallargs;
   cygwin_buf_t *args;
+  char **old_env;
   size_t len;
-  int tty;
+  int tty, i;
   int ostdin, ostdout, ostderr;
 #else
   char real_path[__PMAX];
@@ -2066,6 +2067,8 @@ windows_create_inferior (struct target_o
   strcat (args, cygallargs);
 #endif

+  old_env = environ;
+  environ = in_env;
   /* Prepare the environment vars for CreateProcess.  */
   cygwin_internal (CW_SYNC_WINENV);

@@ -2101,6 +2104,21 @@ windows_create_inferior (struct target_o
                       NULL,    /* current directory */
                       &si,
                       &pi);
+
+  /* Reset all environment variables to avoid leftover on next run. */
+  for (i = 0; in_env[i] && *in_env[i]; i++)
+    {
+      char *equalpos = strchr (in_env[i], '=');
+      if (equalpos)
+       *equalpos = '\0';
+      SetEnvironmentVariableA (in_env[i], NULL);
+      if (equalpos)
+       *equalpos = '=';
+    }
+  /* Restore normal GDB environment variables.  */
+  environ = old_env;
+  cygwin_internal (CW_SYNC_WINENV);
+
   if (tty >= 0)
     {
       close (tty);