Summary: | Tracing options not displayed in Run-time Workbench run configuration | ||
---|---|---|---|
Product: | [Eclipse Project] Platform | Reporter: | Jan Ploski <jpljpl> |
Component: | SWT | Assignee: | Christophe Cornu <christophe.cornu+eclipse> |
Status: | RESOLVED FIXED | QA Contact: | |
Severity: | critical | ||
Priority: | P1 | CC: | cozzi |
Version: | 2.1 | ||
Target Milestone: | --- | ||
Hardware: | PC | ||
OS: | Linux-GTK | ||
Whiteboard: |
Description
Jan Ploski
2003-04-28 10:09:14 EDT
We didn't change one line from 2.1 in PDE space, so I will move this report to PLatform UI first. If this turns out to be a SWT problem, please move further accordingly. org.eclipse.ui.views.properties hasn't changed since 2.1 either. Passing onto swt. Chrix, please get to the bottom of the ASAP. I have not been able to reproduce on RH 8 GTk2.0.6. Which linux distribution, GTK version and locale are you using? (to get the locale: type in locale from a shell). We will try to reproduce it. Thanks, Chris The regression could be in the SWT GTK Table.getItemHeight. The value h[0] should not be <=0 but it seems to be on your machine. public int getItemHeight () { checkWidget(); if (itemCount == 0) { int column = OS.gtk_tree_view_get_column (handle, 0); int list = OS.gtk_tree_view_column_get_cell_renderers (column); int length = OS.g_list_length (list); int renderer = OS.g_list_nth_data (list, length - 1); OS.g_list_free (list); int [] w = new int [1], h = new int [1]; OS.gtk_cell_renderer_get_size (renderer, handle, null, null, null, w, h); if (h[0] <= 0) System.out.println("BUG! "+h[0]); return h [0]; } GdkRectangle rect = new GdkRectangle (); int path = OS.gtk_tree_path_new_first (); OS.gtk_tree_view_get_cell_area (handle, path, 0, rect); OS.gtk_tree_path_free (path); return rect.height; } LANG=C LC_CTYPE="C" LC_NUMERIC="C" LC_TIME="C" LC_COLLATE="C" LC_MONETARY="C" LC_MESSAGES="C" LC_PAPER="C" LC_NAME="C" LC_ADDRESS="C" LC_TELEPHONE="C" LC_MEASUREMENT="C" LC_IDENTIFICATION="C" LC_ALL= GTK 2.2.1-4 from Debian unstable. I can easily reproduce it in the normal, top-level workbench. Let me check if I can also trigger the problem in a run-time workbench in debugging mode. Here are the results of my debugging session: 1. In Table.java:763 itemCount == 1, so it's not entering the branch with h[0] as you supposed. 2. After the call to gtk_tree_view_get_cell_area, rect is still (0, 0, 0, 0). 3. Therefore in TableTreeItem.java:186 itemHeight == 0 4. So the Image constructor fails. Does it help? Anything else I can check? I don't know GTK myself, unfortunately. I just ran the same debugging session under Eclipse 2.1 (on the same PC), which does not throw the exception. The only difference seems to be that in Table.getItemHeight the call to OS.gtk_tree_view_get_cell_area sets a non-zero dimension in the rectangle object. The arguments for the call look much the same as in I20030422 (i.e. the handle and path are both big, positive integers and differ). So I guess you should check if OS.gtk_tree_view_get_cell_area has been touched since 2.1. Ok, I can reproduce the problem with GTK 2.2.1. Raising priority to P1. Fixed in version > 20030429 Thanks Jan for reporting this. *** Bug 37797 has been marked as a duplicate of this bug. *** |