- Code: Select all
browser->GetMainFrame()->LoadRequest(next_request);
- Code: Select all
browser->GetMainFrame()->LoadURL(url);
In both cases, the current displayed page and the next page are from different domains, so the renderer process should not be shared from page to page - at least that's what my understanding of the Chromium process model is at the moment.
I have to use LoadRequest instead of LoadURL.
This creates the problem of memory bloat - the memory used by a previous page is not completely freed in the renderer process and thus accumulates over time.
Is this the intended behavior? How do I work around this?