on the board here.
Works.
Tested-by: Gerd Hoffmann
thanks,
Gerd
___
U-Boot mailing list
U-Boot@lists.denx.de
https://lists.denx.de/listinfo/u-boot
Hi,
> https://www.mail-archive.com/u-boot@lists.denx.de/msg241949.html fixes this
> problem.
Tested things today, only to figure it doesn't work :(
Reverting commit "3f66149 Remove extra fdt_fixup_ethernet() call" brings
things into working state. So the fix above actually works, and
apparent
Hi,
u-boot 2016.11 & newer doesn't set the mac address on my raspberry pi
any more, so the linux kernel picks a random mac address. Bisected to
this commit:
commit b91c6a1209e7da1a7f989d9ac35d0d8be0b7b710
Author: Simon Glass
Date: Wed Oct 5 20:42:11 2016 -0600
Fix return value in trail
Hi,
> Signed-off-by: Gerd Hoffmann
No, I didn't signed this patch off.
And I will not because it is broken.
> for (cnt = (maxsize / sizeof(long)) >> 1; cnt > 0; cnt >>= 1) {
> addr = base + cnt; /* pointer arith! */
> @@ -44,6 +44,7
On Di, 2014-10-21 at 22:14 +0200, Wolfgang Denk wrote:
> Dear Gerd Hoffmann,
>
> In message <1413910153-5907-1-git-send-email-kra...@redhat.com> you wrote:
> > The original memory sizing code in get_ram_size clobbers the word
> > at the base address, but forgets to
stay cleared until the detection is
done, otherwise we'll fail to detect the same piece of memory being
mapped multiple times into the address space.
Cc: Wolfgang Denk
Cc: Iwo Mergler
Signed-off-by: Gerd Hoffmann
---
common/memsize.c | 11 +++
1 file changed, 7 insertions(+), 4
gone (zeroed
out) after calling get_ram_size.
Cc: Wolfgang Denk
Signed-off-by: Gerd Hoffmann
---
common/memsize.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/common/memsize.c b/common/memsize.c
index 589400d..dac1368 100644
--- a/common/memsize.c
+++ b/common/memsize.c
Hi,
> I cannot see any memory corruption. If I understand you correctly,
> the contents of address 0x should be overwritten here?
Depends on the address space layout. vexpress-a9 has flash mapped at
0x. RAM starts at 0x6000, and four bytes at 0x6000 get
corrupted.
Po
On Di, 2014-10-21 at 12:05 +0200, Gerd Hoffmann wrote:
> Check if the fdt is there (happens only when passed via -dtb cmd line
> switch to qemu), then setup env accordingly. If present use it.
> Otherwise try to load from disk.
Oops, wrong patch, correct one just send.
It shows what
calling get_ram_size.
Signed-off-by: Gerd Hoffmann
---
common/memsize.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/common/memsize.c b/common/memsize.c
index 589400d..ad79ff1 100644
--- a/common/memsize.c
+++ b/common/memsize.c
@@ -48,7 +48,9 @@ long get_ram_size(long *base
Check if the fdt is there (happens only when passed via -dtb cmd line
switch to qemu), then setup env accordingly. If present use it.
Otherwise try to load from disk.
Also tweak CONFIG_SYS_LOAD_ADDR and LINUX_BOOT_PARAM_ADDR a bit to
avoid them overriding the fdt.
---
board/armltd/vexpress/vexpr
base[0] is saved, but never restored.
Signed-off-by: Gerd Hoffmann
---
common/memsize.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/common/memsize.c b/common/memsize.c
index 589400d..8bda2ba 100644
--- a/common/memsize.c
+++ b/common/memsize.c
@@ -48,7 +48,9 @@ long
12 matches
Mail list logo
