> > Didn't here anything on this? What is our final decision here? > > It's gotten worse, not better. Apparently, you aren't getting the > concept of protecting the open count with a static lock and my > explanations are just not vivid enough or something. So I decided > to fix it myself. Maybe then the patch in C will explain it better > than English. But I didn't have time to do it.
Probably I'm not trying to do what you want. I analyzed locks for other usb drivers in 2.4 tree and used same ideas. Static lock minor_table_mutex is used for minor table structure. And dev->sem for dev manipulations and that's why for open_count. If you will simply browse /drivers/usb dir for 2.4 you will see that such approach is widely used there. What's not right? Certainly, you have more experience so I can't say that I'm right. > Also, there's an outright bug in the latest version. Your purge > of the wrong lock was incomplete and so there was an unbalanced up(). > But this is moot. Yes, got it. It's up for minor_table_mutex in adu_release. Corrected. > So, the version before the latest is borderline acceptable. If Willy > wants to take it, it's fine. I'll fix it up later together with 2.6. Let's do everything correctly for 2.4. V. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/