[U-Boot-Users] make: *** [u-boot] Error 139
Hi,
I'm trying to compile u-boot version 1.1.2 unmodifed on an iBook G4 with the following specs:
glibc: 2.3.4 gcc: 3.4.3 binutils: 2.15.90.0.3-r3
Here are the steps I'm using, please let me know if I've screwed something up:
make Sandpoint8245_config make
The build appears to go through fine until near the end, where it is trying to use ld with the option:
...-Map u-boot.map -o u-boot
Then I see this every time in the same place:
make: *** [u-boot] Error 139
u-boot.map is empty at this point.
The exact same thing happens when I try to build u-boot using a cross compiler built on an x86 machine for a powerpc 603e cpu. u-boot.map is also empty
Also, I get the exact same error when trying to do the same task on a Pegasos machine running Debian. The only difference on the Pegasos machine is that it actually gets part way through creating u-boot.map.u-boot.map actually contains some data.
Any help with this would be greatly appreciated. My attempts to look the issue up through google have yielded results like "check the hardware, bad ram or disk corruption", or, "it's the fault of binutils 2.15".
Any assistance would be greatly appreciated, as I've wasted two days changing and rebuilding cross compilers, and testing on different machines.
Thanks in advance, Anthony
Dear Anthony,
it has always been a good idea to read or even search (!) a bit in the mailing list archives BEFORE posting.
You should consider this, too.
In message 20050331224744.92468.qmail@web88101.mail.re2.yahoo.com you wrote:
...
make: *** [u-boot] Error 139
Your binutils are broken. File a bug report.
disk corruption", or, "it's the fault of binutils 2.15".
The latter is the correct answer.
Any assistance would be greatly appreciated, as I've wasted two days changing and rebuilding cross compilers, and testing on different machines.
What a waste of effort. The very same problem has been discussed here JUST YESTERDAY.
Best regards,
Wolfgang Denk
Too right :)
On Apr 1, 2005 8:09 AM, Wolfgang Denk wd@denx.de wrote:
Dear Anthony,
it has always been a good idea to read or even search (!) a bit in the mailing list archives BEFORE posting.
You should consider this, too.
In message 20050331224744.92468.qmail@web88101.mail.re2.yahoo.com you wrote:
...
make: *** [u-boot] Error 139
Your binutils are broken. File a bug report.
disk corruption", or, "it's the fault of binutils 2.15".
The latter is the correct answer.
Any assistance would be greatly appreciated, as I've wasted two days changing and rebuilding cross compilers, and testing on different machines.
What a waste of effort. The very same problem has been discussed here JUST YESTERDAY.
Best regards,
Wolfgang Denk
-- Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd@denx.de Linux - Why use Windows, since there is a door? -- Alain Knaff in slrn50c2o5.auk.knaff@choiseul.inrialpes.fr
This SF.net email is sponsored by Demarc: A global provider of Threat Management Solutions. Download our HomeAdmin security software for free today! http://www.demarc.com/info/Sentarus/hamr30 _______________________________________________ U-Boot-Users mailing list U-Boot-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/u-boot-users
Anthony Russello writes:
Anthony> Hi, I'm trying to compile u-boot version 1.1.2 unmodifed on Anthony> an iBook G4 with the following specs:
Anthony> glibc: 2.3.4 gcc: 3.4.3 binutils: 2.15.90.0.3-r3
Anthony> Here are the steps I'm using, please let me know if I've Anthony> screwed something up:
Anthony> make Sandpoint8245_config make
Anthony> The build appears to go through fine until near the end, Anthony> where it is trying to use ld with the option:
Anthony> ...-Map u-boot.map -o u-boot
Anthony> Then I see this every time in the same place:
Anthony> make: *** [u-boot] Error 139
Anthony> u-boot.map is empty at this point.
Anthony> The exact same thing happens when I try to build u-boot Anthony> using a cross compiler built on an x86 machine for a Anthony> powerpc 603e cpu. u-boot.map is also empty
Anthony> Also, I get the exact same error when trying to do the same Anthony> task on a Pegasos machine running Debian. The only Anthony> difference on the Pegasos machine is that it actually gets Anthony> part way through creating u-boot.map.u-boot.map actually Anthony> contains some data.
...
Looks like the problem which I found long time ago in binutils but have no time to fix. What I discovered is that if you request link map (-Map) and the linker command file (u-boot.lds) contains conditional statement (in u-boot it's usually DEFINED(env_offset) ? env_offset : .) then the linker crashes. I don't know if Sandpoint needs this statement, maybe it can be removed, otherwise remove -Map.
In message 16973.32808.124167.511374@home.localdomain you wrote:
Looks like the problem which I found long time ago in binutils but have no time to fix. What I discovered is that if you request link map (-Map) and the linker command file (u-boot.lds) contains conditional statement (in u-boot it's usually DEFINED(env_offset) ? env_offset : .) then the linker crashes. I don't know if Sandpoint needs this statement, maybe it can be removed, otherwise remove -Map.
Ummm... this is definitely NOT the right way to address such a problem.
Instead, the bug in the linker should get fixed. For this is it necessary to report it. So please report it, if necessary again.
Best regards,
Wolfgang Denk
On Apr 2, 2005 11:47 PM, Wolfgang Denk wd@denx.de wrote:
Ummm... this is definitely NOT the right way to address such a problem.
Instead, the bug in the linker should get fixed. For this is it necessary to report it. So please report it, if necessary again.
I've made a report to bug-binutils@gnu.org. Absolutely no clue as to when exactly this will be dealt with, but at least its a step towards right direction.
In message 9b7ca65705040216556848fe4b@mail.gmail.com you wrote:
I've made a report to bug-binutils@gnu.org. Absolutely no clue as to when exactly this will be dealt with, but at least its a step towards right direction.
Thanks. Ummm.. It shouldn't be too difficult to debug given thefact that it's so easily reproducable. How about trying it yourself?
Best regards,
Wolfgang Denk
On Apr 3, 2005 5:03 PM, Wolfgang Denk wd@denx.de wrote:
Thanks. Ummm.. It shouldn't be too difficult to debug given thefact that it's so easily reproducable. How about trying it yourself?
Hmmm that's something I've never thought about doing. May sound like I'm trying to put this aside, but right at this moment, I've got a headache that's been grown to the size of migraine over 1 year period. After this is over, I'll see what I can do to return the favor to the gnu folks. :)
participants (4)
-
Anthony Russello -
Daniel Ann -
Wolfgang Denk -
Yuli Barcohen