From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:34186) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1e4jyl-0005CJ-JT for qemu-devel@nongnu.org; Wed, 18 Oct 2017 04:45:32 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1e4jyi-0007Z8-IQ for qemu-devel@nongnu.org; Wed, 18 Oct 2017 04:45:31 -0400 Received: from mx1.redhat.com ([209.132.183.28]:37706) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1e4jyi-0007YV-Bt for qemu-devel@nongnu.org; Wed, 18 Oct 2017 04:45:28 -0400 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 4BD667EA8D for ; Wed, 18 Oct 2017 08:45:27 +0000 (UTC) From: Juan Quintela In-Reply-To: <20171012055709.GF30901@pxdev.xzpeter.org> (Peter Xu's message of "Thu, 12 Oct 2017 13:57:09 +0800") References: <20171010181542.24168-1-quintela@redhat.com> <20171010181542.24168-7-quintela@redhat.com> <20171012055709.GF30901@pxdev.xzpeter.org> Reply-To: quintela@redhat.com Date: Wed, 18 Oct 2017 10:45:22 +0200 Message-ID: <87d15kondp.fsf@secure.laptop> MIME-Version: 1.0 Content-Type: text/plain Subject: Re: [Qemu-devel] [PATCH 06/10] migration: Make sure that we pass the right cache size List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Xu Cc: qemu-devel@nongnu.org, dgilbert@redhat.com, lvivier@redhat.com Peter Xu wrote: > On Tue, Oct 10, 2017 at 08:15:38PM +0200, Juan Quintela wrote: >> Instead of passing silently round down the number of pages, make it an >> error that the cache size is not a multiple of 2. > > s/multiple/power/? Thanks. > Would this patch break existing users? I have a problem here. Current code: - "silently" truncate the value that we pass - "silently" uses a different value that the one that it shows as used - "shows" a value that is different of what has been set to So, I have to break one of them: - only allow valid values (and then we output the same value that user input) - store "setup" value, and use something different "internaly" No clear winner. This way we are at least consistent with everything else. Documentation already required a power of 2 value. And we give a good error message. Later, Juan. > >> >> Signed-off-by: Juan Quintela >> --- >> migration/page_cache.c | 12 +++++++----- >> 1 file changed, 7 insertions(+), 5 deletions(-) >> >> diff --git a/migration/page_cache.c b/migration/page_cache.c >> index 9a9d13d6a2..96268c3aea 100644 >> --- a/migration/page_cache.c >> +++ b/migration/page_cache.c >> @@ -58,6 +58,13 @@ PageCache *cache_init(int64_t new_size, size_t page_size, Error **errp) >> return NULL; >> } >> >> + /* round down to the nearest power of 2 */ >> + if (!is_power_of_2(num_pages)) { >> + error_setg(errp, QERR_INVALID_PARAMETER_VALUE, "cache size", >> + "is not a power of two number of pages"); >> + return NULL; >> + } >> + >> /* We prefer not to abort if there is no memory */ >> cache = g_try_malloc(sizeof(*cache)); >> if (!cache) { >> @@ -65,11 +72,6 @@ PageCache *cache_init(int64_t new_size, size_t page_size, Error **errp) >> "Failed to allocate cache"); >> return NULL; >> } >> - /* round down to the nearest power of 2 */ >> - if (!is_power_of_2(num_pages)) { >> - num_pages = pow2floor(num_pages); >> - DPRINTF("rounding down to %" PRId64 "\n", num_pages); >> - } >> cache->page_size = page_size; >> cache->num_items = 0; >> cache->max_num_items = num_pages; >> -- >> 2.13.6 >>