diff mbox

[u-boot,v2,1/3] aspeed: remove hack loading ramdisk in memory

Message ID 1480717624-24994-2-git-send-email-raltherr@google.com
State Superseded, archived
Headers show

Commit Message

Rick Altherr Dec. 2, 2016, 10:27 p.m. UTC 
From: Cédric Le Goater <clg@kaod.org>

Signed-off-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Rick Altherr <raltherr@google.com>
---
 common/image.c | 11 -----------
 1 file changed, 11 deletions(-)

Comments

Simon Glass Dec. 5, 2016, 6:24 a.m. UTC | #1 
On 2 December 2016 at 15:27, Rick Altherr <raltherr@google.com> wrote:
> From: Cédric Le Goater <clg@kaod.org>
>
> Signed-off-by: Cédric Le Goater <clg@kaod.org>
> Signed-off-by: Rick Altherr <raltherr@google.com>
> ---
>  common/image.c | 11 -----------
>  1 file changed, 11 deletions(-)

Reviewed-by: Simon Glass <sjg@chromium.org>
Cédric Le Goater Dec. 5, 2016, 8:13 a.m. UTC | #2 
Hello,

On 12/05/2016 07:24 AM, Simon Glass wrote:
> On 2 December 2016 at 15:27, Rick Altherr <raltherr@google.com> wrote:
>> From: Cédric Le Goater <clg@kaod.org>
>>
>> Signed-off-by: Cédric Le Goater <clg@kaod.org>
>> Signed-off-by: Rick Altherr <raltherr@google.com>
>> ---
>>  common/image.c | 11 -----------
>>  1 file changed, 11 deletions(-)
> 
> Reviewed-by: Simon Glass <sjg@chromium.org>
> 

So how do you fix the ramdisk loading ? is using the FIT images 
enough to work around this problem ? 

For legacy type images, this patch seemed to work : 

	https://github.com/legoater/u-boot/commit/86fd073a12650ee7fc4e6514f2b0f159be275f29

I am not sure if this is a good approach though.

Thanks,

C.
Rick Altherr Dec. 5, 2016, 5:09 p.m. UTC | #3 
See the next patch in this series.  OpenBMC uses ATAGS when building a
legacy image.  U-Boot ATAGS support wasn't causing ramdisk relocation.
It's not really a FIT vs legacy issue so much as a FDT vs ATAGS issue.  We
could still use legacy images with FDT and the ramdisk would be relocated
fine.  I fixed the ATAGS support to trigger ramdisk relocation as that's
the actual bug.

On Mon, Dec 5, 2016 at 12:13 AM, Cédric Le Goater <clg@kaod.org> wrote:

> Hello,
>
> On 12/05/2016 07:24 AM, Simon Glass wrote:
> > On 2 December 2016 at 15:27, Rick Altherr <raltherr@google.com> wrote:
> >> From: Cédric Le Goater <clg@kaod.org>
> >>
> >> Signed-off-by: Cédric Le Goater <clg@kaod.org>
> >> Signed-off-by: Rick Altherr <raltherr@google.com>
> >> ---
> >>  common/image.c | 11 -----------
> >>  1 file changed, 11 deletions(-)
> >
> > Reviewed-by: Simon Glass <sjg@chromium.org>
> >
>
> So how do you fix the ramdisk loading ? is using the FIT images
> enough to work around this problem ?
>
> For legacy type images, this patch seemed to work :
>
>         https://github.com/legoater/u-boot/commit/
> 86fd073a12650ee7fc4e6514f2b0f159be275f29
>
> I am not sure if this is a good approach though.
>
> Thanks,
>
> C.
>
diff mbox

Patch

diff --git a/common/image.c b/common/image.c
index 20c4807..0be09e5 100644
--- a/common/image.c
+++ b/common/image.c
@@ -1069,17 +1069,6 @@  int boot_get_ramdisk(int argc, char * const argv[], bootm_headers_t *images,
 				return 1;
 			}
 		}
-/// TODO ... Check Why ..................
-#if  defined(CONFIG_ARCH_ASPEED)
-		/*
-		 * We need to copy the ramdisk to SRAM to let Linux boot
-		 */
-		if (rd_data) {
-			memmove ((void *)rd_load, (uchar *)rd_data, rd_len);
-			rd_data = rd_load;
-		}
-#endif /* CONFIG_ASPEED */
-
 	} else if (images->legacy_hdr_valid &&
 			image_check_type(&images->legacy_hdr_os_copy,
 						IH_TYPE_MULTI)) {