qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Haozhong Zhang <haozhong.zhang@intel.com>
To: Eduardo Habkost <ehabkost@redhat.com>
Cc: qemu-devel@nongnu.org, Igor Mammedov <imammedo@redhat.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Peter Crosthwaite <crosthwaite.peter@gmail.com>,
	Richard Henderson <rth@twiddle.net>
Subject: Re: [Qemu-devel] [PATCH v2 3/3] hostmem-file: make option 'size' optional
Date: Wed, 2 Nov 2016 10:08:33 +0800	[thread overview]
Message-ID: <20161102020833.lkpaxqpmliehlcr6@hz-desktop> (raw)
In-Reply-To: <20161031181834.GA25480@thinpad.lan.raisama.net>

On 10/31/16 16:18 -0200, Eduardo Habkost wrote:
>On Fri, Oct 28, 2016 at 10:06:40AM +0800, Haozhong Zhang wrote:
>[...]
>> > > diff --git a/exec.c b/exec.c
>> > > index 264a25f..89065bd 100644
>> > > --- a/exec.c
>> > > +++ b/exec.c
>> > > @@ -1234,7 +1234,7 @@ static int64_t get_file_size(int fd)
>> > >  }
>> > >
>> > >  static void *file_ram_alloc(RAMBlock *block,
>> > > -                            ram_addr_t memory,
>> > > +                            ram_addr_t *memory,
>> > >                              const char *path,
>> > >                              Error **errp)
>> > >  {
>> > > @@ -1245,6 +1245,7 @@ static void *file_ram_alloc(RAMBlock *block,
>> > >      void *area = MAP_FAILED;
>> > >      int fd = -1;
>> > >      int64_t file_size;
>> > > +    ram_addr_t mem_size = *memory;
>> > >
>> > >      if (kvm_enabled() && !kvm_has_sync_mmu()) {
>> > >          error_setg(errp,
>> > > @@ -1309,21 +1310,27 @@ static void *file_ram_alloc(RAMBlock *block,
>> > >
>> > >      file_size = get_file_size(fd);
>> > >
>> > > -    if (memory < block->page_size) {
>> > > +    if (!mem_size && file_size > 0) {
>> > > +        mem_size = file_size;
>> >
>> > Maybe we should set *memory here and not below?
>> >
>>
>> Qemu currently sets the memory region size to the file size, and block
>> length to the aligned file size, so the code here can be changed as below:
>>
>>            memory_region_set_size(block->mr, mem_size);
>>            mem_size = HOST_PAGE_ALIGN(mem_size);
>>            *memory = mem_size;
>>
>> The second line is necessary because Qemu currently passes the aligned
>> file size to file_ram_alloc().
>
>That would duplicate the existing HOST_PAGE_ALIGN logic from
>qemu_ram_alloc_from_file(), won't it?
>
>I believe that's yet another reason to check file size before
>initializing the memory region, instead of initializing it first,
>and fixing up its size later.
>

Agree. In the next version of this patch 3, I'll move file operations
(open/close/get size) from file_ram_alloc() to qemu_ram_alloc_from_file().

Thanks,
Haozhong

  reply	other threads:[~2016-11-02  2:08 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-10-27  4:22 [Qemu-devel] [PATCH v2 0/3] Improve truncation behavior of memory-backend-file Haozhong Zhang
2016-10-27  4:22 ` [Qemu-devel] [PATCH v2 1/3] exec.c: do not truncate non-empty memory backend file Haozhong Zhang
2016-10-27 14:31   ` Eduardo Habkost
2016-10-28  2:07     ` Haozhong Zhang
2016-10-31 17:21       ` Eduardo Habkost
2016-10-27  4:22 ` [Qemu-devel] [PATCH v2 2/3] exec.c: check memory backend file size with 'size' option Haozhong Zhang
2016-10-27 14:32   ` Eduardo Habkost
2016-10-31 17:23   ` Eduardo Habkost
2016-10-31 17:56     ` Paolo Bonzini
2016-11-02  1:05   ` [Qemu-devel] [RESEND PATCH " Haozhong Zhang
2016-10-27  4:23 ` [Qemu-devel] [PATCH v2 3/3] hostmem-file: make option 'size' optional Haozhong Zhang
2016-10-27 14:55   ` Eduardo Habkost
2016-10-28  2:06     ` Haozhong Zhang
2016-10-28  5:57       ` Haozhong Zhang
2016-10-31 18:18       ` Eduardo Habkost
2016-11-02  2:08         ` Haozhong Zhang [this message]
2016-10-27 12:03 ` [Qemu-devel] [PATCH v2 0/3] Improve truncation behavior of memory-backend-file Paolo Bonzini

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161102020833.lkpaxqpmliehlcr6@hz-desktop \
    --to=haozhong.zhang@intel.com \
    --cc=crosthwaite.peter@gmail.com \
    --cc=ehabkost@redhat.com \
    --cc=imammedo@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).