Bug 1969533

Summary: Eclipse SWT Tree leaks native memory via gtk_tree_view_expand_row()
Product: Red Hat Enterprise Linux 7 Reporter: Simeon Andreev <simeon.andreev>
Component: gtk3Assignee: Matthias Clasen <mclasen>
Status: CLOSED MIGRATED QA Contact: Desktop QE <desktop-qa-list>
Severity: high Docs Contact:
Priority: medium    
Version: 7.9CC: akurtako, loskutov, mcoufal, otte, sbarcomb, vikpatil, vkadlcik
Target Milestone: rcKeywords: MigratedToJIRA
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2023-09-15 16:56:04 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Simeon Andreev 2021-06-08 14:40:19 UTC
Description of problem:

Eclipse SWT Tree leaks native memory via gtk_tree_view_expand_row()


Version-Release number of selected component (if applicable):

We observe the leak with Eclipse 4.15, Eclipse 4.20 and Eclipse 4.21, as well as latest master from SWT. We assume the leak is seen with all Eclipse versions inbetween.

How reproducible:

To reproduce the leak, run the following SWT snippet:

public static void main(String[] args) {
    Display display = new Display();
    Shell shell = new Shell(display);
    shell.setText("Tree Example");
    shell.setLayout(new FillLayout());
    Tree tree = new Tree(shell, SWT.NONE);
    tree.setItemCount(2);
    TreeItem item0 = tree.getItem(0);
    item0.setText("item0");
    TreeItem item1 = tree.getItem(1);
    item1.setText("item1");
    item1.setItemCount(6);
    TreeItem[] children = item1.getItems();
    for (int k = 0; k < children.length; ++k) {
        children[k].setText("child" + k);
    }
    item1.setExpanded(true);
    tree.setItemCount(1);
    shell.open();
    while (!shell.isDisposed()) {
        if (!display.readAndDispatch())
            display.sleep();
    }
    display.dispose();
}

jemalloc reports a memory leak as follows:

[240 bytes leaked]
je_prof_backtrace (/home/sandreev/git/misc/jemalloc/src/prof.c:636 (discriminator 2))
je_malloc_default (/home/sandreev/git/misc/jemalloc/src/jemalloc.c:2289)
g_malloc (??:?)
g_slice_alloc (??:?)
_gtk_rbnode_new.isra.2 (/usr/src/debug/gtk+-3.22.30/gtk/gtkrbtree.c:61)
_gtk_rbtree_insert_after (/usr/src/debug/gtk+-3.22.30/gtk/gtkrbtree.c:456)
gtk_tree_view_build_tree (/usr/src/debug/gtk+-3.22.30/gtk/gtktreeview.c:9570)
gtk_tree_view_real_expand_row (/usr/src/debug/gtk+-3.22.30/gtk/gtktreeview.c:12865)
gtk_tree_view_expand_row (/usr/src/debug/gtk+-3.22.30/gtk/gtktreeview.c:12920)
Java_org_eclipse_swt_internal_gtk_GTK_gtk_1tree_1view_1expand_1row (??:?)
?? (??:0)

The same leak can be reproduced with the following GTK+ snippet:

// gcc -g tree.c `pkg-config --cflags --libs gtk+-3.0` -o tree

#include <gtk/gtk.h>

enum
{
  COL1 = 0,
  NUM_COLS
} ;

int
main (int argc, char **argv)
{
  GtkWidget *window;
  GtkWidget *scrolled_window;
  GtkWidget *view;

  gtk_init (&argc, &argv);

  window = gtk_window_new (GTK_WINDOW_TOPLEVEL);
  g_signal_connect (window, "delete_event", gtk_main_quit, NULL);

  gtk_window_set_default_size(GTK_WINDOW(window), 200, 200);

  scrolled_window = gtk_scrolled_window_new(NULL, NULL);

  GtkCellRenderer     *renderer;

  view = gtk_tree_view_new ();

  renderer = gtk_cell_renderer_text_new ();
  gtk_tree_view_insert_column_with_attributes (GTK_TREE_VIEW (view), -1, "column 1", renderer, "text", COL1, NULL);

  GtkTreeStore *treestore;
  GtkTreeIter   iter, child;

  treestore = gtk_tree_store_new(NUM_COLS, G_TYPE_STRING);

  gtk_tree_store_append(treestore, &iter, NULL);
  gtk_tree_store_set(treestore, &iter, COL1, "item", -1);
  gtk_tree_store_append(treestore, &iter, NULL);
  gtk_tree_store_set(treestore, &iter, COL1, "item", -1);
  GtkTreePath *path = NULL;
  path = gtk_tree_model_get_path (GTK_TREE_MODEL(treestore), &iter);

  int j;
  for (j = 0; j < 6; ++j)
  {
    gtk_tree_store_append(treestore, &child, &iter);
    gtk_tree_store_set(treestore, &child, COL1, "child", -1);
  }

  gtk_tree_view_set_model (GTK_TREE_VIEW (view), GTK_TREE_MODEL (treestore));

  if (path)
  {
    gtk_tree_view_expand_row(GTK_TREE_VIEW(view), path, FALSE);
    gtk_tree_path_free(path);
  }

  gtk_container_add (GTK_CONTAINER (window), scrolled_window);
  gtk_container_add (GTK_CONTAINER (scrolled_window), view);

  gtk_widget_show_all (window);

  GtkTreeIter iter1, child1;
  if (gtk_tree_model_iter_nth_child(GTK_TREE_MODEL(treestore), &iter1, NULL, 1))
  {
    gtk_tree_store_remove(treestore, &iter1);
  }

  g_object_unref(treestore);

  gtk_main ();

  return 0;
}


jemalloc reports a memory leak as follows:

[240 bytes leaked]
je_prof_backtrace (/home/sandreev/git/misc/jemalloc/src/prof.c:636 (discriminator 2))
je_malloc_default (/home/sandreev/git/misc/jemalloc/src/jemalloc.c:2289)
g_malloc (??:?)
g_slice_alloc (??:?)
_gtk_rbnode_new.isra.2 (/usr/src/debug/gtk+-3.22.30/gtk/gtkrbtree.c:61)
_gtk_rbtree_insert_after (/usr/src/debug/gtk+-3.22.30/gtk/gtkrbtree.c:456)
gtk_tree_view_build_tree (/usr/src/debug/gtk+-3.22.30/gtk/gtktreeview.c:9570)
gtk_tree_view_real_expand_row (/usr/src/debug/gtk+-3.22.30/gtk/gtktreeview.c:12865)
gtk_tree_view_expand_row (/usr/src/debug/gtk+-3.22.30/gtk/gtktreeview.c:12920)
?? (??:0)
__libc_start_main (/usr/src/debug/glibc-2.17-c758a686/csu/../csu/libc-start.c:274)
?? (??:0)


Actual results:

Memory is leaked.

Expected results:

Memory is not leaked.

Additional info:

For more context (e.g. actual Eclipse user operations to run into the leak) see Eclipse bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=574065

Comment 2 Simeon Andreev 2021-06-08 14:42:20 UTC
We need either a GTK+  developer or a developer familiar with GTK3 API usage to look into this and let us know if:

1. This is a memory leak in GTK+.
1.1. In this case we need a fix in GTK+, which we require to resolve this support case.
2. SWT is supposed to use different API to remove tree rows, to avoid leaking memory.
2.1. We need to know what API we are supposed to use, to remove tree rows without causing a memory leak.
2.2. We need to know what to expect of the new code performance-wise. E.g. should we expect massive performance degradation?

RHEL support case: https://access.redhat.com/support/cases/#/case/02959748

Comment 3 Andrey Loskutov 2021-06-08 14:54:16 UTC
Simeon, please change title. It is not Eclipse nor SWT that leaks memory here, but GTK library via gtk_tree_view_expand_row(). Also the "component" is wrong, it must be GTK.

Comment 4 Simeon Andreev 2021-06-08 14:59:24 UTC
(In reply to Andrey Loskutov from comment #3)
> Simeon, please change title. It is not Eclipse nor SWT that leaks memory
> here, but GTK library via gtk_tree_view_expand_row(). Also the "component"
> is wrong, it must be GTK.

We don't know whether SWT is misusing GTK+ APIs to cause the memory leak, see comment 2.

Comment 8 RHEL Program Management 2023-09-15 16:52:06 UTC
Issue migration from Bugzilla to Jira is in process at this time. This will be the last message in Jira copied from the Bugzilla bug.

Comment 9 RHEL Program Management 2023-09-15 16:56:04 UTC
This BZ has been automatically migrated to the issues.redhat.com Red Hat Issue Tracker. All future work related to this report will be managed there.

Due to differences in account names between systems, some fields were not replicated.  Be sure to add yourself to Jira issue's "Watchers" field to continue receiving updates and add others to the "Need Info From" field to continue requesting information.

To find the migrated issue, look in the "Links" section for a direct link to the new issue location. The issue key will have an icon of 2 footprints next to it, and begin with "RHEL-" followed by an integer.  You can also find this issue by visiting https://issues.redhat.com/issues/?jql= and searching the "Bugzilla Bug" field for this BZ's number, e.g. a search like:

"Bugzilla Bug" = 1234567

In the event you have trouble locating or viewing this issue, you can file an issue by sending mail to rh-issues. You can also visit https://access.redhat.com/articles/7032570 for general account information.