8 Feb
2012
8 Feb
'12
1:32 p.m.
I've pulled git repo today, but the RSSI firmware gets an error.
apps/rssi/main.c: In function `main':
apps/rssi/main.c:896: warning: 'a' might be used uninitialized in this
function
apps/rssi/main.c:896: warning: 'e' might be used uninitialized in this
function
CC board/compal_e88/rssi.compalram.manifest.o
LD board/compal_e88/rssi.compalram.elf
OBJ board/compal_e88/rssi.compalram.bin
CC board/compal_e88/rssi.highram.manifest.o
LD board/compal_e88/rssi.highram.elf
OBJ board/compal_e88/rssi.highram.bin
CC board/compal_e88/rssi.e88loader.manifest.o
LD board/compal_e88/rssi.e88loader.elf
OBJ board/compal_e88/rssi.e88loader.bin
CC board/compal_e88/rssi.e88flash.manifest.o
LD board/compal_e88/rssi.e88flash.elf
OBJ board/compal_e88/rssi.e88flash.bin
CC board/compal_e86/rssi.compalram.manifest.o
LD board/compal_e86/rssi.compalram.elf
arm-elf-ld: region LRAM is full (board/compal_e86/rssi.compalram.elf
section .data)
make[1]: *** [board/compal_e86/rssi.compalram.elf] Error 1
make[1]: Leaving directory src/target/firmware'
make: *** [firmware] Error 2
$ git pull
Already up-to-date.
$
Anyone experiencing the same issue?
9 Feb
9 Feb
7:41 a.m.
Dario Lombardo wrote
arm-elf-ld: region LRAM is full (board/compal_e86/rssi.compalram.elf
section .data)
make[1]: *** [board/compal_e86/rssi.compalram.elf] Error 1
make[1]: Leaving directory src/target/firmware'
make: *** [firmware] Error 2
Anyone experiencing the same issue?
I've got the same error
(if TX is enabled)
--
View this message in context:
http://baseband-devel.722152.n3.nabble.com/Can-t-compile-after-last-pull-tp…
Sent from the baseband-devel mailing list archive at Nabble.com.
1:44 p.m.
Hi,
On 09.02.2012 07:41, User wrote:
> Anyone experiencing the same issue?
It seems your toolchain produces a binary larger than 80 kBytes
(0x14000), which is the current limit for compalram images.
What we need is a setting in the Makefile to blacklist certain apps
from being built for certain environments, but what you can do for now
is increase the LRAM length and adjust the IRAM ORIGIN in firmware/board
/compal/ram.lds.
Regards,
Steve
2:56 p.m.
Steve Markgraf wrote:
Hi,
On 09.02.2012 07:41, User wrote:
would it be possible to make a linker script that dynamically sets code
and data memory size?
> Anyone experiencing the same issue?
>
It seems your toolchain produces a binary larger than 80 kBytes
(0x14000), which is the current limit for compalram images.
What we need is a setting in the Makefile to blacklist certain apps
from being built for certain environments, but what you can do for now
is increase the LRAM length and adjust the IRAM ORIGIN in firmware/board
/compal/ram.lds.
Regards,
Steve
10 Feb
10 Feb
10:06 a.m.
I've tried this:
MEMORY
{
/* compal-loaded binary: our text, initialized data */
LRAM (rw) : ORIGIN = 0x00800000, LENGTH = 0x00015000
/* compal-loaded binary: our unitialized data, stacks, heap */
IRAM (rw) : ORIGIN = 0x00815000, LENGTH = 0x0000c000
}
but
LD board/compal_e88/rssi.compalram.elf
OBJ board/compal_e88/rssi.compalram.bin
LD board/compal_e86/rssi.compalram.elf
OBJ board/compal_e86/rssi.compalram.bin
LD board/compal_e86/rssi.highram.elf
arm-elf-ld: region LRAM is full (board/compal_e86/rssi.highram.elf section
.data)
arm-elf-ld: section .got [00820000 -> 00820013] overlaps section
.text.start [00820000 -> 0082006b]
make: *** [board/compal_e86/rssi.highram.elf] Error 1
I have the same behaviour also with other values of length/origin.
Any other suggestion?
On Thu, Feb 9, 2012 at 1:44 PM, Steve Markgraf <steve(a)steve-m.de> wrote:
Hi,
On 09.02.2012 07:41, User wrote:
> Anyone experiencing the same issue?
It seems your toolchain produces a binary larger than 80 kBytes
(0x14000), which is the current limit for compalram images.
What we need is a setting in the Makefile to blacklist certain apps
from being built for certain environments, but what you can do for now
is increase the LRAM length and adjust the IRAM ORIGIN in firmware/board
/compal/ram.lds.
Regards,
Steve
13 Feb
13 Feb
5:46 a.m.
New subject: rare warning?
make[1]: warning: Clock skew detected. Your build may be incomplete.
CygWin git.February
thank you
13 Feb
13 Feb
12:30 a.m.
New subject: rare warning?
thank you
----- Original Message -----
From: "Peter Stuge" <peter(a)stuge.se>
To: <baseband-devel(a)lists.osmocom.org>
Sent: Tuesday, February 14, 2012 2:19 AM
Subject: Re: rare warning?
Juantxi wrote:
make[1]: warning: Clock skew detected. Your build
may be incomplete.
Your system clock has been adjusted backwards. Some files have dates
in the future.
//Peter
9:06 a.m.
Now it works.
I've tried this:
I changed ram.lds & highram.lds.
ram.lds:
LRAM (rw) : ORIGIN = 0x00800000, LENGTH = 0x00015000
IRAM (rw) : ORIGIN = 0x00815000, LENGTH = 0x0000c000
highram.lds:
LRAM (rw) : ORIGIN = 0x00820000, LENGTH = 0x00015000
IRAM (rw) : ORIGIN = 0x00835000, LENGTH = 0x0000c000
--
View this message in context:
http://baseband-devel.722152.n3.nabble.com/Can-t-compile-after-last-pull-tp…
Sent from the baseband-devel mailing list archive at Nabble.com.
17 Dec
17 Dec
2:12 p.m.
This also worked for me. As a beginner, I remind everybody to edit the files
in "compal", not the ram.lds in "mediatek" like I first did after
"find".
--
View this message in context:
http://baseband-devel.722152.n3.nabble.com/Can-t-compile-after-last-pull-tp…
Sent from the baseband-devel mailing list archive at Nabble.com.
2:14 p.m.
Uh, my reply was to the post with the LRAM and IRAM addresses. No hierarchy
in threading on this forum?
--
View this message in context:
http://baseband-devel.722152.n3.nabble.com/Can-t-compile-after-last-pull-tp…
Sent from the baseband-devel mailing list archive at Nabble.com.
20 Dec
20 Dec
11:26 p.m.
Uh, my reply was to the post with the LRAM and IRAM
addresses. No hierarchy
in threading on this forum?
It's not a forum, it's a mailing list ...
And you should use sylvain/testing branch, it contains a revised build
system that should work without modifications
Cheers,
Sylvain
4152
days inactive
4468
days old
baseband-devel@lists.osmocom.org
11 comments
9 participants
participants (9)
-
- -
Andreas Eversberg -
Anssi Porttikivi -
Dario Lombardo -
Juantxi -
Peter Stuge -
Steve Markgraf -
Sylvain Munaut -
User