From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:34492) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XgcGV-0004Rg-JT for qemu-devel@nongnu.org; Tue, 21 Oct 2014 12:26:37 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1XgcGP-0003M5-EF for qemu-devel@nongnu.org; Tue, 21 Oct 2014 12:26:31 -0400 Received: from mx1.redhat.com ([209.132.183.28]:50854) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XgcGP-0003Lp-6b for qemu-devel@nongnu.org; Tue, 21 Oct 2014 12:26:25 -0400 Message-ID: <5446892B.1050109@redhat.com> Date: Tue, 21 Oct 2014 18:26:19 +0200 From: Max Reitz MIME-Version: 1.0 References: <1409348463-16627-1-git-send-email-mreitz@redhat.com> <1409348463-16627-2-git-send-email-mreitz@redhat.com> <20141020142510.GM3585@noname.redhat.com> <54451E8E.70402@redhat.com> <20141020144805.GP3585@noname.redhat.com> In-Reply-To: <20141020144805.GP3585@noname.redhat.com> Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH v5 01/11] qcow2: Calculate refcount block entry count List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Kevin Wolf Cc: Stefan Hajnoczi , qemu-devel@nongnu.org, =?windows-1252?Q?Beno=EEt_Canet?= On 2014-10-20 at 16:48, Kevin Wolf wrote: > Am 20.10.2014 um 16:39 hat Max Reitz geschrieben: >> On 20.10.2014 at 16:25, Kevin Wolf wrote: >>> Am 29.08.2014 um 23:40 hat Max Reitz geschrieben: >>>> The size of a refblock entry is (in theory) variable; calculate >>>> therefore the number of entries per refblock and the according bit shift >>>> (1 << x == entry count) when opening an image. >>>> >>>> Signed-off-by: Max Reitz >>>> --- >>>> block/qcow2.c | 2 ++ >>>> block/qcow2.h | 2 ++ >>>> 2 files changed, 4 insertions(+) >>>> >>>> diff --git a/block/qcow2.c b/block/qcow2.c >>>> index f9e045f..172ad00 100644 >>>> --- a/block/qcow2.c >>>> +++ b/block/qcow2.c >>>> @@ -689,6 +689,8 @@ static int qcow2_open(BlockDriverState *bs, QDict *options, int flags, >>>> s->l2_bits = s->cluster_bits - 3; /* L2 is always one cluster */ >>>> s->l2_size = 1 << s->l2_bits; >>>> + s->refcount_block_bits = s->cluster_bits - (s->refcount_order - 3); >>>> + s->refcount_block_size = 1 << s->refcount_block_bits; >>>> bs->total_sectors = header.size / 512; >>>> s->csize_shift = (62 - (s->cluster_bits - 8)); >>>> s->csize_mask = (1 << (s->cluster_bits - 8)) - 1; >>>> diff --git a/block/qcow2.h b/block/qcow2.h >>>> index 6aeb7ea..7c01fb7 100644 >>>> --- a/block/qcow2.h >>>> +++ b/block/qcow2.h >>>> @@ -222,6 +222,8 @@ typedef struct BDRVQcowState { >>>> int l2_size; >>>> int l1_size; >>>> int l1_vm_state_index; >>>> + int refcount_block_bits; >>>> + int refcount_block_size; >>> Might just be me, but size sounds to me as if the unit were bytes. Would >>> you mind renaming this as refcount_block_entries or refblock_entries? >> If I'm doing a v7, no. Otherwise, well, see l1_size and l2_size. ;-) > Good point. This has confused people more than once, so it's probably > not just me. Okay, now that I've done it and was about to send the series and just wanted to convert a local variable named refcount_table_size to refcount_table_entries, I decided not do do this. I'll call it refcount_block_size in v7 as well. The reason for this is that I started looking for "_size" in block/qcow2-refcount.c. "_entries" is never used, the number of entries per L1, L2 or refcount table is always foo_size. In BDRVQcowState, there's not only l1_size and l2_size, but refcount_table_size as well. Calling it refcount_block_entries without renaming those would be extremely weird, and renaming those does not seem like a viable option to me. Furthermore, I'd find it extremely confusing to have "_entries" in some places and "_size" in others when there's no difference between the two. Currently, people ask "Why is this foo_size an entry count? ... Well, okay, that seems to be just the way it is." With foo_entries, it'll be "Why is this foo_size an entry count when bar_entries exists, so shouldn't it be foo_entries if they want it to be an entry count?" I'll keep it as refcount_block_size, although I'm afraid reverting all these changes will be as hard as having made them in the first place... Oh, well, here goes. Max