All of lore.kernel.org
 help / color / mirror / Atom feed
From: Fawad Lateef <fawadlateef@gmail.com>
To: Zachary Amsden <zach@vmware.com>
Cc: linux-kernel@vger.kernel.org
Subject: Re: Highmemory Problem with RHEL3 .... 2.4.21-5.ELsmp
Date: Mon, 8 Aug 2005 12:32:12 +0500	[thread overview]
Message-ID: <1e62d13705080800329114fdd@mail.gmail.com> (raw)
In-Reply-To: <42F6EB73.3030104@vmware.com>

On 8/8/05, Zachary Amsden <zach@vmware.com> wrote:
> 
> IIRC 2.4.21 has some highmem bugs that have since been fixed.  But, it
> sounds like you might be doing something quite unusual.  Code would
> definitely give people a better idea of what might be wrong.  

I think you overlooked what i mentioned in P.S. ; which is 

My memory reservation and later using that memory through kmap_atomic
works well on the kernels other than RHEL3 2.4.21-e.ELsmp
.............. the page reservation was done in the
arch/i386/mm/init.c file in function one_highpage_init ...... I have
Machine with 16GB RAM and 2 - Xeon 2.4GHz Processors .....

The code which I added for memory reservation in kernel is : 

void __init one_highpage_init(struct page *page, int pfn, int bad_ppro)
{
	if (!page_is_ram(pfn)) {
		SetPageReserved(page);
		return;
	}
	
	if (bad_ppro && page_kills_ppro(pfn)) {
		SetPageReserved(page);
		return;
	}

// Here's the code which i added for memory reservation ..... i m
setting 0xC4 in page->count just because i will know later that these
pages have been reserved by me ... not by kernel .....

	if ((unsigned long)(page - mem_map) > 0x80000) {
		SetPageReserved(page);
		set_bit(PG_highmem, &page->flags);
		atomic_set(&page->count, 0xC4);
		totalhigh_pages++;
		return;
	}

// My code Ends here 
	
	ClearPageReserved(page);
	set_bit(PG_highmem, &page->flags);
	atomic_set(&page->count, 1);
	__free_page(page);
	totalhigh_pages++;
}


After this in my module, i simply use kmap_atomic to map the page
reserved by me and tried to use that ........ its working perfect in
both 2.4.x series and also working in 2.6.x .....

> You should definitely consider moving to 2.6 to get a better response.
> 

i already moved to 2.6.x already !!! but the current requiment is to
use RHEL3 Kernel which is 2.4.21-27.ELsmp

I think its now more clear !!!! waiting for your resposes !!!


-- 
Fawad Lateef

  reply	other threads:[~2005-08-08  7:32 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-08-08  3:50 Highmemory Problem with RHEL3 .... 2.4.21-5.ELsmp Fawad Lateef
2005-08-08  4:34 ` Nish Aravamudan
2005-08-08  5:19 ` Zachary Amsden
2005-08-08  7:32   ` Fawad Lateef [this message]
2005-08-08  7:42     ` Zachary Amsden
2005-08-08  8:17 ` Arjan van de Ven
2005-08-08 17:24   ` Fawad Lateef
2005-08-08 17:44     ` Arjan van de Ven

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=1e62d13705080800329114fdd@mail.gmail.com \
    --to=fawadlateef@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=zach@vmware.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.