Le 15/02/13 15:34, Fredrik Orderud a écrit :

On Tue, Feb 12, 2013 at 7:08 PM, Vicente J. Botet Escriba <vicente.botet@wanadoo.fr> wrote:

I will start to analyze your patch this week.

I don't know if you have followed this ticket.

#7906

Very bad performance of generic implementation of shared_mutex on windows

From the performances presented it seems that the windows specific implementation will be needed, so maybe a windows specific implementation for shared priority mutex should also be needed. The problem is that the windows implementation contains some bugs and missing features.

Thank you for the update Vicente. I was not aware of the poor win32 performance for generic shared_mutex, so that of course needs to be taken into consideration. I believe it should be feasible to refactor the pthread & win32 shared_mutex implementations to share state_data (locking policy logic) even though the locking "backends" are different.

yes, but I don't know if a generic backend for priority shared mutex will be enough for windows, so if you have some ideas on how to implements them on windows it is worth exploring them. Anyway, it will be better to have a generic implementation of priority shared mutex than none.

I would like to introduce your patch soon, but lastly there has been a lot of troubles and I have not be tempted to add more disturbance.

I'll let you know when I have some news.