[Issue 3454] New: Inconsistent flag setting in GC.realloc()

Oct 30, 2009

David Simcha

Oct 25, 2013

safety0ff.bugz

Oct 28, 2013

safety0ff.bugz

Oct 31, 2013

github-bugzilla@puremagic.com

October 30, 2009

[Issue 3454] New: Inconsistent flag setting in GC.realloc()

Posted by David Simcha

Permalink

David Simcha

Permalink

http://d.puremagic.com/issues/show_bug.cgi?id=3454

           Summary: Inconsistent flag setting in GC.realloc()
           Product: D
           Version: 2.035
          Platform: Other
        OS/Version: Windows
            Status: NEW
          Severity: normal
          Priority: P2
         Component: druntime
        AssignedTo: sean@invisibleduck.org
        ReportedBy: dsimcha@yahoo.com


--- Comment #0 from David Simcha <dsimcha@yahoo.com> 2009-10-29 17:26:42 PDT ---
GC.realloc() doesn't set the flags on either the old or new memory block for a whole bunch of code paths.  I was tipped off to this by reading the code while trying to add precise heap scanning.  Here's a test program that demonstrates this.

import std.stdio, core.memory;

void main() {
    doTest(1);
    writeln();
    doTest(1024 * 1024);
}

void doTest(size_t multiplier) {
    auto foo = GC.malloc(8 * multiplier);
    auto bar = GC.realloc(foo, 2 * multiplier, GC.BlkAttr.NO_SCAN);

    writeln("Old block attributes:  ", GC.getAttr(foo));
    writeln("New block attributes:  ", GC.getAttr(bar));
    writeln("Old Ptr:  ", foo, "  New Ptr:  ", bar);
}

Output:
Old block attributes:  2
New block attributes:  2
Old Ptr:  961E40  New Ptr:  961E30

Old block attributes:  0
New block attributes:  0
Old Ptr:  10C0000  New Ptr:  10C0000


This is caused by the GC returning early from the B_PAGE path, or if the new block is almost the same size as the old block.  If I get precise heap scanning to work, I'll include a fix for this in the patch.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=3454 safety0ff.bugz <safety0ff.bugz@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |safety0ff.bugz@gmail.com --- Comment #1 from safety0ff.bugz <safety0ff.bugz@gmail.com> 2013-10-25 12:07:56 PDT --- I believe your test should be: ------------------------------------------------ import std.stdio, core.memory; void main() { doTest(1); writeln(); doTest(1024 * 1024); } void doTest(size_t multiplier) { auto foo = GC.malloc(8 * multiplier); writeln("Old block attributes: ", GC.getAttr(foo)); auto bar = GC.realloc(foo, 2 * multiplier, GC.BlkAttr.NO_SCAN); writeln("New block attributes: ", GC.getAttr(bar)); writeln("Old Ptr: ", foo, " New Ptr: ", bar); } ------------------------------------------------- I've put an ugly fix for this here: https://github.com/Safety0ff/druntime/commit/14148c8184b7094243f5ab74d703027d05c6e73a -- Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email ------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=3454 --- Comment #3 from github-bugzilla@puremagic.com 2013-10-31 10:26:34 PDT --- Commits pushed to master at https://github.com/D-Programming-Language/druntime https://github.com/D-Programming-Language/druntime/commit/4fe17c32bcfa9140e3b52e92b27474c3b60b3f95 fix Issue 3454 - inconsistent flag setting in GC.realloc() https://github.com/D-Programming-Language/druntime/commit/487f0a1ed1aad76d1b4ca638fd25b730fd87aa4f regression test for Issue 3454 https://github.com/D-Programming-Language/druntime/commit/0fe338f42a1225a6041bb0f7707375cfbee7b54b Merge pull request #645 from Safety0ff/fix3454 fix Issue 3454 - inconsistent flag setting in GC.realloc() -- Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email ------- You are receiving this mail because: -------

Forums