All of lore.kernel.org
 help / color / mirror / Atom feed
From: Stefan Althoefer <stefan.althoefer@web.de>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH] IXP425: make cmd_go handle caches correctly Part 1/1
Date: Wed, 17 Dec 2008 00:11:08 +0100	[thread overview]
Message-ID: <gi9cfv$224$1@ger.gmane.org> (raw)
In-Reply-To: <20081215231824.EE9F8832E8A1@gemini.denx.de>

IXP425: make cmd_go handle caches correctly

This patch adds icache_invalidate() and dcache_flush() to
cpu/ixp/cpu.c.

Also it adds do_go_exec() which is called from do_go().
This private implementation performs cache handling
before jumping into the code.

Without this cache handling, you will jump into stale
code if you download a program several times.

Signed-off-by: Stefan Althoefer <stefan.althoefer@web.de>
---

>
> Coding style: please use TABs for indentation.
>

Sorry for the spaces.

BTW: In function calls: Is the space between function name
are opening bracket desired or not? Seems this is handled
very inconsistent.

-- Stefan

 cpu/ixp/cpu.c |   32 ++++++++++++++++++++++++++++++++
 1 files changed, 32 insertions(+), 0 deletions(-)

diff --git a/cpu/ixp/cpu.c b/cpu/ixp/cpu.c
index 27872fb..9035d98 100644
--- a/cpu/ixp/cpu.c
+++ b/cpu/ixp/cpu.c
@@ -171,6 +171,22 @@ int icache_status (void)
 	return (i & 0x1000);
 }

+void icache_invalidate (void)
+{
+	register u32 i;
+
+	/* FIXME: Required to check if icache is enabled before we do this? */
+
+	/* invalidate I-cache (i is ignored)*/
+	i = 0;
+	asm ("mcr p15, 0, %0, c7, c5, 0": :"r" (i));
+
+	/* cpwait */
+	asm ("mrc  p15,0,%0,c2,c0,0;\
+	    mov  %0,%0;\
+	    sub  pc,pc,#4" : : "r" (i) );
+}
+
 /* we will never enable dcache, because we have to setup MMU first */
 void dcache_enable (void)
 {
@@ -187,6 +203,11 @@ int dcache_status (void)
 	return 0;					/* always off */
 }

+void dcache_flush (void)
+{
+	return;
+}
+
 /* FIXME */
 /*
 void pci_init(void)
@@ -195,6 +216,17 @@ void pci_init(void)
 }
 */

+/* With IXP (and ARM generally) we need to handle caches to avoid
+   stale data. */
+unsigned long do_go_exec (ulong (*entry)(int, char *[]), int argc, char *argv[])
+{
+	dcache_flush ();
+	icache_invalidate ();
+
+	return entry (argc, argv);
+}
+
+
 #ifdef CONFIG_BOOTCOUNT_LIMIT

 void bootcount_store (ulong a)
-- 
1.5.6

  reply	other threads:[~2008-12-16 23:11 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-12-04 21:09 [U-Boot] [PATCH] IXP425: make cmd_go handle caches correctly Part 1/1 Stefan Althoefer
2008-12-15 23:18 ` Wolfgang Denk
2008-12-16 23:11   ` Stefan Althoefer [this message]
2009-01-04  7:51     ` Jean-Christophe PLAGNIOL-VILLARD

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='gi9cfv$224$1@ger.gmane.org' \
    --to=stefan.althoefer@web.de \
    --cc=u-boot@lists.denx.de \
    /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.