LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Alexander van Heukelum <heukelum@mailshack.com>
To: Ingo Molnar <mingo@elte.hu>
Cc: Alexander van Heukelum <heukelum@fastmail.fm>,
	Ian Campbell <ijc@hellion.org.uk>,
	Mark McLoughlin <markmc@redhat.com>,
	"H. Peter Anvin" <hpa@zytor.com>, Andi Kleen <ak@suse.de>,
	Thomas Gleixner <tglx@linutronix.de>,
	Jeremy Fitzhardinge <jeremy@goop.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] reserve end-of-conventional-memory to 1MB on 64-bit add-on
Date: Sat, 1 Mar 2008 17:12:43 +0100	[thread overview]
Message-ID: <20080301161243.GA13301@mailshack.com> (raw)
In-Reply-To: <20080301160911.GA13271@mailshack.com>

This patch is an add-on to the 64-bit ebda patch. It makes
the functions reserve_ebda_region (renamed from reserve_ebda)
and copy_e820_map equal to the 32-bit versions of the previous
patch.

Changes:

Use u64 and u32 for local variables in copy_e820_map.

The amount of conventional memory and the start of the EBDA are
detected by reading the BIOS data area directly. Paravirtual
environments do not provide this area, so we bail out early
in that case. They will just have to set up a correct memory
map to start with.

Add a safety net for zeroed out BIOS data area.

Signed-off-by: Alexander van Heukelum <heukelum@fastmail.fm>

---

diff --git a/arch/x86/kernel/e820_32.c b/arch/x86/kernel/e820_32.c
diff --git a/arch/x86/kernel/e820_64.c b/arch/x86/kernel/e820_64.c
index dd68cfd..baa4223 100644
--- a/arch/x86/kernel/e820_64.c
+++ b/arch/x86/kernel/e820_64.c
@@ -621,10 +621,10 @@ static int __init copy_e820_map(struct e820entry *biosmap, int nr_map)
 		return -1;
 
 	do {
-		unsigned long start = biosmap->addr;
-		unsigned long size = biosmap->size;
-		unsigned long end = start + size;
-		unsigned long type = biosmap->type;
+		u64 start = biosmap->addr;
+		u64 size = biosmap->size;
+		u64 end = start + size;
+		u32 type = biosmap->type;
 
 		/* Overflow in 64 bits? Ignore the memory map. */
 		if (start > end)
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index b684552..269a6b4 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -55,12 +55,30 @@ static void __init copy_bootdata(char *real_mode_data)
 /*
  * The BIOS places the EBDA/XBDA at the top of conventional
  * memory, and usually decreases the reported amount of
- * conventional memory (int 0x12) too.
+ * conventional memory (int 0x12) too. This also contains a
+ * workaround for Dell systems that neglect to reserve EBDA.
+ * The same workaround also avoids a problem with the AMD768MPX
+ * chipset: reserve a page before VGA to prevent PCI prefetch
+ * into it (errata #56). Usually the page is reserved anyways,
+ * unless you have no PS/2 mouse plugged in.
  */
-static __init void reserve_ebda(void)
+static void __init reserve_ebda_region(void)
 {
 	unsigned int lowmem, ebda_addr;
 
+	/* To determine the position of the EBDA and the */
+	/* end of conventional memory, we need to look at */
+	/* the BIOS data area. In a paravirtual environment */
+	/* that area is absent. We'll just have to assume */
+	/* that the paravirt case can handle memory setup */
+	/* correctly, without our help. */
+#ifdef CONFIG_PARAVIRT
+	if ((boot_params.hdr.version >= 0x207) &&
+			(boot_params.hdr.hardware_subarch != 0)) {
+		return;
+	}
+#endif
+
 	/* end of low (conventional) memory */
 	lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES);
 	lowmem <<= 10;
@@ -80,8 +98,8 @@ static __init void reserve_ebda(void)
 		lowmem = 0x9f000;
 
 	/* Paranoia: should never happen, but... */
-	if (lowmem >= 0x100000)
-		lowmem = 0xa0000;
+	if ((lowmem == 0) || (lowmem >= 0x100000))
+		lowmem = 0x9f000;
 
 	/* reserve all memory between lowmem and the 1MB mark */
 	reserve_early(lowmem, 0x100000, "BIOS reserved");
@@ -140,7 +158,7 @@ void __init x86_64_start_kernel(char * real_mode_data)
 		reserve_early(ramdisk_image, ramdisk_end, "RAMDISK");
 	}
 
-	reserve_ebda();
+	reserve_ebda_region();
 
 	/*
 	 * At this point everything still needed from the boot loader
diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c

  reply	other threads:[~2008-03-01 16:14 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-02-24 17:46 [PATCH] Fix alignment of early reservation for EBDA Alexander van Heukelum
2008-02-24 19:27 ` Andi Kleen
2008-02-24 19:41 ` Ingo Molnar
2008-02-24 20:53   ` Alexander van Heukelum
2008-02-25  2:18 ` H. Peter Anvin
2008-02-25 16:54   ` Alexander van Heukelum
2008-02-25 17:01     ` Ingo Molnar
2008-02-25 18:07       ` [PATCH] reserve_early end-of-conventional-memory to 1MB Alexander van Heukelum
2008-02-25 18:13         ` H. Peter Anvin
2008-02-25 19:46           ` Alexander van Heukelum
2008-02-25 21:17             ` H. Peter Anvin
2008-02-26  9:30             ` Ingo Molnar
2008-02-27 14:26               ` Andi Kleen
2008-02-27 14:38               ` [PATCH] reserve_early end-of-conventional-memory to 1MB II - some numbers to put it into perspective Andi Kleen
2008-02-27 16:44                 ` H. Peter Anvin
2008-02-27 20:01                 ` Alexander van Heukelum
2008-02-28 13:13         ` [PATCH] reserve end-of-conventional-memory to 1MB on 32-bit Alexander van Heukelum
2008-02-28 13:28           ` [RFC] use realmode code to reserve end-of-conventional-memory to 1MB Alexander van Heukelum
2008-02-28 21:12             ` Ian Campbell
2008-02-28 21:14               ` H. Peter Anvin
2008-02-28 23:16                 ` Ian Campbell
2008-02-29 20:00                   ` Ingo Molnar
2008-03-04 11:41                   ` Mark McLoughlin
2008-03-04 14:33                     ` Ingo Molnar
2008-03-04 15:12                       ` Ian Campbell
2008-03-04 15:13                       ` Jeremy Fitzhardinge
2008-03-04 15:25                         ` Ingo Molnar
2008-03-04 16:02                           ` Jeremy Fitzhardinge
2008-03-04 16:15                             ` Ingo Molnar
2008-03-04 16:15                       ` Mark McLoughlin
2008-03-04 16:23                         ` Ingo Molnar
2008-03-04 17:44                   ` Jeremy Fitzhardinge
2008-03-05 15:59                   ` Eduardo Habkost
2008-03-05 16:08                     ` H. Peter Anvin
2008-03-05 16:53                       ` Eduardo Habkost
2008-03-05 17:28                         ` H. Peter Anvin
2008-03-05 17:28                           ` Jeremy Fitzhardinge
2008-03-05 17:38                             ` H. Peter Anvin
2008-03-05 16:38                     ` Jeremy Fitzhardinge
2008-03-05 17:27                       ` H. Peter Anvin
2008-02-28 21:09           ` [PATCH] reserve end-of-conventional-memory to 1MB on 32-bit Ian Campbell
2008-02-29 11:49             ` Alexander van Heukelum
2008-02-29 17:14               ` Mark McLoughlin
2008-02-29 18:38                 ` Alexander van Heukelum
2008-02-29 18:44                   ` H. Peter Anvin
2008-02-29 18:56                     ` Alexander van Heukelum
2008-02-29 22:06                   ` Mark McLoughlin
2008-02-29 22:26                     ` Alexander van Heukelum
2008-03-01 16:09                     ` [PATCH] reserve end-of-conventional-memory to 1MB on 32-bit v2 Alexander van Heukelum
2008-03-01 16:12                       ` Alexander van Heukelum [this message]
2008-03-04 11:44                       ` Mark McLoughlin
2008-03-04 13:31                         ` Alexander van Heukelum
2008-03-04 14:49                         ` Ingo Molnar
2008-03-04 15:16                           ` Mark McLoughlin
2008-03-04 15:24                             ` Ingo Molnar
2008-03-04 15:18                         ` Jeremy Fitzhardinge
2008-03-04 16:51                           ` Alexander van Heukelum
2008-03-04 17:05                             ` H. Peter Anvin
2008-03-04 17:11                             ` Jeremy Fitzhardinge
2008-03-04 18:57                               ` [PATCH] reserve end-of-conventional-memory to 1MB, 32-bit, use paravirt_enabled Alexander van Heukelum
2008-03-04 19:12                                 ` [PATCH] reserve end-of-conventional-memory to 1MB, 64-bit, " Alexander van Heukelum
2008-02-27 14:25     ` [PATCH] Fix alignment of early reservation for EBDA Andi Kleen

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=20080301161243.GA13301@mailshack.com \
    --to=heukelum@mailshack.com \
    --cc=ak@suse.de \
    --cc=heukelum@fastmail.fm \
    --cc=hpa@zytor.com \
    --cc=ijc@hellion.org.uk \
    --cc=jeremy@goop.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=markmc@redhat.com \
    --cc=mingo@elte.hu \
    --cc=tglx@linutronix.de \
    --subject='Re: [PATCH] reserve end-of-conventional-memory to 1MB on 64-bit add-on' \
    /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

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).