substring locations and # line directives (PR preprocessor/78569)

The ICE in PR preprocessor/78569 appears to be due to an attempt to
generate substring locations in a .i file where the underlying .c file
has changed since the .i file was generated.

This can't work, so it seems safest for the on-demand substring
locations to be unavailable for such files, falling back to
"whole string" locations for such cases.

gcc/ChangeLog:
	PR preprocessor/78569
	* input.c (get_substring_ranges_for_loc): Fail gracefully if
	line directives were present.

gcc/testsuite/ChangeLog:
	PR preprocessor/78569
	* gcc.dg/format/pr78569.c: New test case.

From-SVN: r242990
This commit is contained in:
David Malcolm 2016-11-30 01:13:37 +00:00 committed by David Malcolm
parent 3517d13be7
commit 94f597df22
4 changed files with 45 additions and 0 deletions

View file

@ -1,3 +1,9 @@
2016-11-29 David Malcolm <dmalcolm@redhat.com>
PR preprocessor/78569
* input.c (get_substring_ranges_for_loc): Fail gracefully if
line directives were present.
2016-11-30 Matthias Klose <doko@ubuntu.com>
* doc/install.texi: Document configure options --enable-objc-gc

View file

@ -1331,6 +1331,16 @@ get_substring_ranges_for_loc (cpp_reader *pfile,
if (cpp_get_options (pfile)->track_macro_expansion != 2)
return "track_macro_expansion != 2";
/* If #line or # 44 "file"-style directives are present, then there's
no guarantee that the line numbers we have can be used to locate
the strings. For example, we might have a .i file with # directives
pointing back to lines within a .c file, but the .c file might
have been edited since the .i file was created.
In such a case, the safest course is to disable on-demand substring
locations. */
if (line_table->seen_line_directive)
return "seen line directive";
/* If string concatenation has occurred at STRLOC, get the locations
of all of the literal tokens making up the compound string.
Otherwise, just use STRLOC. */

View file

@ -1,3 +1,8 @@
2016-11-29 David Malcolm <dmalcolm@redhat.com>
PR preprocessor/78569
* gcc.dg/format/pr78569.c: New test case.
2016-11-29 Martin Sebor <msebor@redhat.com>
avoid false positives on ILP32 targets.

View file

@ -0,0 +1,24 @@
/* { dg-do compile } */
/* { dg-options "-Wformat-length" } */
/* A run of blank lines, so that we would fail the assertion in input.c:1388:
gcc_assert (line_width >= (start.column - 1 + literal_length)); */
void test (void)
{
char tmp[128];
/* Point to the run of blank lines, so that the components of the overlong
string appear to be present within the run of blank lines. */
# 6 "../../../../src/gcc/testsuite/gcc.dg/format/pr78569.c"
__builtin_snprintf (tmp, sizeof(tmp),
"The Base Band sends this value as a response to a "
"request for IMSI detach sent over the control "
"channel uplink (see section 7.6.1).");
/* { dg-warning "output truncated" "" { target *-*-* } 7 } */
/* { dg-message "format output" "" { target *-*-* } 6 } */
}