From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtpng2.m.smailru.net (smtpng2.m.smailru.net [94.100.179.3]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dev.tarantool.org (Postfix) with ESMTPS id 15FF246971A for ; Fri, 13 Dec 2019 16:37:23 +0300 (MSK) Date: Fri, 13 Dec 2019 16:35:14 +0300 From: Igor Munkin Message-ID: <20191213133514.GN1214@tarantool.org> References: <20191210083258.GD21413@atlas> <2c8fe897-9a9d-849d-463e-5fadff982b8c@tarantool.org> <20191211070830.GA5953@atlas> <20191212000047.GL1214@tarantool.org> <1d40cedb-c631-a420-2391-1f27a5a07fe5@tarantool.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <1d40cedb-c631-a420-2391-1f27a5a07fe5@tarantool.org> Subject: Re: [Tarantool-patches] [PATCH 2/2] fiber: destroy fiber.storage created by iproto List-Id: Tarantool development patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Vladislav Shpilevoy Cc: tarantool-patches@dev.tarantool.org Vlad, On 13.12.19, Vladislav Shpilevoy wrote: > Thanks for the review! > > > > > >> /* reset pending wakeups */ > >> rlist_del(&fiber->state); > >> if (! (fiber->flags & FIBER_IS_JOINABLE)) > >> diff --git a/src/lib/core/fiber.h b/src/lib/core/fiber.h > >> index c5b975513..21fff8f88 100644 > >> --- a/src/lib/core/fiber.h > >> +++ b/src/lib/core/fiber.h > >> @@ -458,8 +458,17 @@ struct fiber { > >> > >> /** Triggers invoked before this fiber yields. Must not throw. */ > >> struct rlist on_yield; > >> - /** Triggers invoked before this fiber stops. Must not throw. */ > >> - struct rlist on_stop; > >> + /** > >> + * Triggers invoked before this fiber is > >> + * stopped/reset/recycled/destroyed/reused. In other > >> + * words, each time when the fiber has finished execution > >> + * of a request. > >> + * In particular, for fibers not from fiber pool the > >> + * cleanup is done before destroy and death. > >> + * Pooled fibers are cleaned up after each request, even > >> + * if they are never destroyed. > >> + */ > >> + struct rlist on_cleanup; > > > > Minor: both trigger lists above has the "Must not throw" note. Does the > > introduced list respect this rule? Please, adjust the comment if yes. > > Nope. These 'must not throw' are outdated. Our triggers does not > throw errors anymore by design. So I deleted it from the updated > comment. > OK, thanks, now it's clear. Guess we need an activity to update comments for other trigger lists. -- Best regards, IM