DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 8F32F78523
To: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc: LKML <linux-kernel@vger.kernel.org>,
        =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= <rkrcmar@redhat.com>
From: Denys Vlasenko <dvlasenk@redhat.com>
Subject: Can avic_vm_id_bitmap be made smaller? 2 mbytes feels too large
Message-ID: <238915be-22f4-4ea3-addf-0e06700f3c1e@redhat.com>
Date: Wed, 21 Jun 2017 20:28:48 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101
 Thunderbird/45.5.1
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 614
Lines: 17

#define AVIC_VM_ID_BITS                        24
#define AVIC_VM_ID_NR                  (1 << AVIC_VM_ID_BITS)
...
static DECLARE_BITMAP(avic_vm_id_bitmap, AVIC_VM_ID_NR);

The above results in a data object which is 2 megabytes large.

avic_vm_init() -> avic_get_next_vm_id() allocates a new vm_id
by looking for a free bit there. That's the whole purpose
for that bitmap existing.

Is there a way to do this less wastefully? Say, such as
iterating over all existing VMs and picking an id which is
not taken.

I imagine VM creation operation is not that frequent, no need
to make new vm_id selection super fast.