| To: | ananth@xxxxxxx (Rajagopal Ananthanarayanan) |
|---|---|
| Subject: | Re: one more question... |
| From: | mostek@xxxxxxx |
| Date: | Wed, 7 Jun 2000 07:47:53 -0500 (CDT) |
| Cc: | dxm@xxxxxxxxxxxxxxxxxxxxxxxx (Daniel Moore), linux-xfs@xxxxxxxxxxx |
| In-reply-to: | <393F40A7.AAB70E06@xxxxxxx> from "Rajagopal Ananthanarayanan" at Jun 07, 2000 11:43:51 PM |
| Sender: | owner-linux-xfs@xxxxxxxxxxx |
O_SYNC definitely implies no delay_alloc.
O_SYNC forces the write all the way to disk before returning
which means allocation must occur.
Jim
>
>Daniel Moore wrote:
>>
>> Just want to check this one is correct...
>> O_SYNC implies no delayed alloc, doesn't it?
>>
>> page_buf.c:
>>
>> 3307c3307
>> < if (delay_alloc)
>> ---
>> > if (delay_alloc && !sync)
>
>
>Hmm. You seem to have an older version of page_buf.c
>I couldn't see the above code anywhere in tot page_buf.c ...
>
>--------------------------------------------------------------------------
>Rajagopal Ananthanarayanan ("ananth")
>Member Technical Staff, SGI.
>--------------------------------------------------------------------------
>
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: one more question..., Rajagopal Ananthanarayanan |
|---|---|
| Next by Date: | Re: setting device on RT files, Steve Lord |
| Previous by Thread: | Re: one more question..., Rajagopal Ananthanarayanan |
| Next by Thread: | Re: ongoing discussions on linux-mm, Rajagopal Ananthanarayanan |
| Indexes: | [Date] [Thread] [Top] [All Lists] |