U++ forum: Welcome to the forum

Re: TheIDE stuck in infinite loop [message #61257 is a reply to message #61242]

Mon, 16 December 2024 20:29

Didier
Messages: 736
Registered: November 2008
Location: France

Contributor

jjacksonRIAB wrote on Sat, 14 December 2024 18:43

Didier,

Yes, I think you found it. That looks like the same backtrace I had in gdb. It's entirely possible I mistakenly thought clearing the clipboard externally fixed it when it was simply the timeout being reached at coincidentally the same time, although IIRC I have triggered it also during a copy+paste operation. When I tried to paste the contents externally to see what was in there the buffer appeared to be empty. Maybe it had something bad in there?

But mainly what I don't understand is why you could be trying to just type things normally, no special characters or commands and it's checking the clipboard. Is TheIDE doing this after every key event? There may be a good reason for this, but I can't think of one.

I'm gonna lookup the 'key' value : since 'InsertChar(key, count, true)' doesn't react as expected, I suppose it's due to the valey of 'key'.

Context: I was trying to type " virtual void Skin(..."
'v' ==> 0x76
'V' ==> 0x56
'tab' ==> 0x09

The 'key' values cought from the backtrace:
67108873 => 0x0400 0009 looks like 'tab'
69206102 => 0x0420 0056 looks like 'V'

So I guess the problem will comme from the 0x0400 0000
The only thing that I can find that is close to it is:
CtrlCoreFlags::K_KEYUP = 0x4000000

Looking at 'InsertChar(...)' the constants used have the following value on may Linux.
K_CHAR_LIM = 0x0200 000
K_ENTER = 0x0d
K_SHIFT_SPACE = 0x0080 0020

So with the 'key' values from backtrace

                               0xXXX      1       true
bool LineEdit::InsertChar(dword key, int count, bool canow) {
	if(key == K_TAB && !processtab)    **** TEST IS FALSE ****
		return false;
	if(filter && key >= 32 && key < K_CHAR_LIM)   **** TEST IS FALSE ****
		key = (*filter)(key);
	if(!IsReadOnly() && (key >= 32 && key < K_CHAR_LIM || key == '\t' || key == '\n' ||
	   key == K_ENTER && processenter || key == K_SHIFT_SPACE))      **** TEST IS FALSE ****
       {
		if(key >= 128 && key < K_CHAR_LIM && (charset != CHARSET_UTF8 && charset != CHARSET_UTF8_BOM)
		   && FromUnicode((wchar)key, charset) == DEFAULTCHAR)
			return true;
		if(!RemoveSelection() && overwrite && key != '\n' && key != K_ENTER && canow) {
			int64 q = cursor;
			int i = GetLinePos64(q);
			if(q + count - 1 < GetLineLength(i))
				Remove((int)cursor, (int)count);
		}
		WString text(key == K_ENTER ? '\n' : key == K_SHIFT_SPACE ? ' ' : key, count);
		Insert((int)cursor, text, true);
		PlaceCaret(cursor + count);
		Action();
		return true;
	}
	return false;    **** Exit comes from here ****
}

The origin of the problem is definitly the 0x4000000 bit set in the value

Looking at the void Ctrl::Proc() in GtkEvent.cpp ==> CtrlCoreFlags::K_KEYUP is definitly used

I thoink the 0x0400 0000 value comes from the line :

DispatchKey(!pressed * K_KEYUP + kv, CurrentEvent.count);

==> Keyboard is malfunctionning : key returns up before being completly processed ??
Note: I do have a brand new Dell wireless keyboard ==> Is this it : a keyboard problem (or kbd driver problem)????

[Updated on: Mon, 16 December 2024 20:49]

Report message to a moderator