https://bugs.kde.org/show_bug.cgi?id=482137
--- Comment #11 from Marcel Hasler <mahas...@gmail.com> --- (In reply to Fabio from comment #10) > (In reply to Marcel Hasler from comment #9) > > But here's a simple example to show that you cannot rely on the data > > remaining valid: > > I know that when the local function variable goes out of scope it's not safe > to use anymore. > FreeRDP creates an internal duplicate of the "path" string using _strdup(): > > > if ((!isPath && !isSpecial) || !(drive->Path = _strdup(path))) > > From that point FreeRDP will use its internal copy of the string stored in > drive->Path. > So, even if our variable gets deleted, it doesn't matter. The point I was trying to make is that by the time freerdp actually calls _strdup, both temporary QByteArray instances have already been destroyed and the pointers are therefore no longer valid. It just *happens* to work because the data hasn't yet been freed or reused/modified in the meantime, but that doesn't mean the code is correct or that it will continue to work in the future. The data just *happens* to still be available, but really you are relying on undocumented behavior here. In fact, the Qt folks could modify their internal implementation at any point (even from a patch version to the next), causing your code to crash or misbehave in other ways. As I see it, the solution is really simple here, just create two local variables and avoid any potential trouble in the future. -- You are receiving this mail because: You are watching all bug changes.
