Hi Marcin, > I think that the last_with_datap comment needs some adjustments. It > states: > > "The last_with_datap pointer points at _whatever 'next' pointer_ > points at the last_with_datap chain." > > There is probably a typo and it should be "last_with_data chain", not > "last_with_datap chain". > > And I think it would be grammatically correct and clearer if it were: > > "The last_with_datap pointer points at _whatever 'next' pointer_ > pointing at the last_with_data chain."
Indeed, patch? > BTW, I was also looking for (\*.*last_with_datap.*[^=]=[^=]) an > explanation for the pointer's indirectness and found only one occurrence > where .next of some chain is modified via that pointer: > > PRESERVE_PINNED(): *src->last_with_datap = tmp; > > All other cases seem to dereference .last_with_datap to a pointer to an > actual chain first. I don't see "why?", since there is an assert(pinned == last_with_datap) and evbuffer_chain_new() does not change last_with_datap, so it should be changed. Yeah it is confusing. *********************************************************************** To unsubscribe, send an e-mail to [email protected] with unsubscribe libevent-users in the body.
