Author Topic: Symbols browser issue of CC has been fixed for my Linux  (Read 12451 times)

Offline AZ

  • Almost regular
  • **
  • Posts: 163
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #60 on: April 19, 2020, 12:48:52 am »
build.log
CB: nightly build .
Thread model: posix
gcc version 5.1.1 20150618 (Red Hat 5.1.1-4) (GCC)

Offline AZ

  • Almost regular
  • **
  • Posts: 163
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #61 on: April 19, 2020, 12:56:47 am »
now , if i use a previously working version for FC30. I still do not get the window. But in console i observe:
Code: [Select]
(codeblocks:126954): Gtk-CRITICAL **: 18:52:31.070: gtk_tree_model_iter_nth_child: assertion 'n >= 0' failed
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it
18:52:48: Debug: window wxScintilla(0x56207fc408a0, ) lost focus even though it didn't have it



-- terminating the app and i see the below
18:53:22: Debug: 3 threads were not terminated by the application.
CB: nightly build .
Thread model: posix
gcc version 5.1.1 20150618 (Red Hat 5.1.1-4) (GCC)

Offline oBFusCATed

  • Developer
  • Lives here!
  • *****
  • Posts: 12911
    • Travis build status
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #62 on: April 19, 2020, 10:51:45 am »
If the application is crashing you have to use a debugger to see where.
(most of the time I ignore long posts)
[strangers don't send me private messages, I'll ignore them; post a topic in the forum, but first read the rules!]

Offline AZ

  • Almost regular
  • **
  • Posts: 163
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #63 on: April 19, 2020, 10:58:44 pm »
If the application is crashing you have to use a debugger to see where.

It is not terminating. I was terminating it ( alt-F4) . Bad wording.
CB: nightly build .
Thread model: posix
gcc version 5.1.1 20150618 (Red Hat 5.1.1-4) (GCC)

Offline AZ

  • Almost regular
  • **
  • Posts: 163
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #64 on: May 14, 2020, 11:41:55 pm »
the symbol browser "hid" in the "management" window. //don't ask  ::)

The browser has been working perfectly fine for my basic needs ( 5/20 files, up to 200 lines of code in each), for the past 2 months with CB been compiled weekly ( from nighties) 
CB: nightly build .
Thread model: posix
gcc version 5.1.1 20150618 (Red Hat 5.1.1-4) (GCC)

Offline ChristophMS

  • Single posting newcomer
  • *
  • Posts: 1
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #65 on: October 22, 2020, 01:32:55 pm »
Hi folks,

I applied that patch from above to my CB::20.03 (on M$ Windows) and at first I got an segmentation fault in void ClassBrowserBuilderThread::SelectItem(CCTreeCtrl* tree, wxTreeItemId item ) where data->m_MirrorNode is accessed. The address of m_MirrorNode looked strange to me (sth. like 0xbnnn nnnn) which seemed to be a kind of random address. I found out that this class member is not initialized in the constructor which should now look like

CCTreeCtrlData::CCTreeCtrlData(SpecialFolder sf, Token* token, short int kindMask, int parentIdx) :
    m_Token(token),
    m_KindMask(kindMask),
    m_SpecialFolder(sf),
    m_TokenIndex(token ? token->m_Index : -1),
    m_TokenKind(token ? token->m_TokenKind : tkUndefined),
    m_TokenName(token ? token->m_Name : _T("")),
    m_ParentIndex(parentIdx),
    m_Ticket(token ? token->GetTicket() : 0),
    m_MirrorNode( nullptr )

{
}

Kind regards

Christoph MS

Offline Miguel Gimenez

  • Lives here!
  • ****
  • Posts: 574
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #66 on: October 22, 2020, 02:56:54 pm »
Thank you for testing. I have modified the patch to include your change.

EDIT: patch removed, the changes for classbrowser.h were missing due to an edition error. I will publish soon a fully threaded symbol explorer patch.
« Last Edit: November 04, 2020, 10:29:39 am by Miguel Gimenez »

Offline Miguel Gimenez

  • Lives here!
  • ****
  • Posts: 574
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #67 on: November 11, 2020, 09:23:06 am »
I have just created a ticket with the new fully threaded implementation, patch is there.

https://sourceforge.net/p/codeblocks/tickets/1031/

Offline tigerbeard

  • Multiple posting newcomer
  • *
  • Posts: 105
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #68 on: November 11, 2020, 10:22:44 am »
I have just created a ticket with the new fully threaded implementation, patch is there.
Thanks a lot for your effort!

Offline oBFusCATed

  • Developer
  • Lives here!
  • *****
  • Posts: 12911
    • Travis build status
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #69 on: November 11, 2020, 03:45:56 pm »
Now people need to test it and report if it works.
(most of the time I ignore long posts)
[strangers don't send me private messages, I'll ignore them; post a topic in the forum, but first read the rules!]

Offline ollydbg

  • Developer
  • Lives here!
  • *****
  • Posts: 5279
  • OpenCV and Robotics
    • Chinese OpenCV forum moderator
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #70 on: November 11, 2020, 03:59:48 pm »
I have just created a ticket with the new fully threaded implementation, patch is there.

https://sourceforge.net/p/codeblocks/tickets/1031/

Good work! I will test it.
If some piece of memory should be reused, turn them to variables (or const variables).
If some piece of operations should be reused, turn them to functions.
If they happened together, then turn them to classes.

Offline ollydbg

  • Developer
  • Lives here!
  • *****
  • Posts: 5279
  • OpenCV and Robotics
    • Chinese OpenCV forum moderator
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #71 on: November 22, 2020, 09:55:23 am »
I have release a C::B windows 64bit pre-release version here:

https://github.com/asmwarrior/codeblocks_sf/releases/tag/pre-release-2020-11-22

So, this patch can be tested by more people.
If some piece of memory should be reused, turn them to variables (or const variables).
If some piece of operations should be reused, turn them to functions.
If they happened together, then turn them to classes.

Offline Miguel Gimenez

  • Lives here!
  • ****
  • Posts: 574
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #72 on: November 22, 2020, 02:37:57 pm »
Does the patch work correctly for you?

Offline ollydbg

  • Developer
  • Lives here!
  • *****
  • Posts: 5279
  • OpenCV and Robotics
    • Chinese OpenCV forum moderator
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #73 on: November 22, 2020, 03:27:01 pm »
Does the patch work correctly for you?
Yes.
If some piece of memory should be reused, turn them to variables (or const variables).
If some piece of operations should be reused, turn them to functions.
If they happened together, then turn them to classes.

Offline oBFusCATed

  • Developer
  • Lives here!
  • *****
  • Posts: 12911
    • Travis build status
Re: Symbols browser issue of CC has been fixed for my Linux
« Reply #74 on: November 22, 2020, 07:36:54 pm »
What does pre-release means? Why don't you just call it "custom version"?
(most of the time I ignore long posts)
[strangers don't send me private messages, I'll ignore them; post a topic in the forum, but first read the rules!]