Compiling Zephyr - At my wits end

IRIX/Nekoware development, porting and related topics.
Forum rules
Any posts concerning pirated software or offering to buy/sell/trade commercial software are subject to removal.
User avatar
praetor242
Posts: 265
Joined: Thu Feb 23, 2017 5:41 pm
Location: Arlington, TX

Compiling Zephyr - At my wits end

Unread postby praetor242 » Tue Nov 14, 2017 5:11 pm

I'm trying to compile MIT's Zephyr IM platform, and it gives me the error message below. I've successfully compiled it's dependency, libss, which works fine (I think). I hate this error message because it doesn't tell you anything, and I've read the man page for SGI's linker with not clues how to track down what's wrong.

ld32: WARNING 84 : /usr/lib/../lib32/libw.so is not used for resolving any symbol.
ld32: ERROR 33 : Unresolved data symbol "_et_list" -- 1st referenced by /usr/nekoware/lib/libss.a(ss_err.o).
Use linker option -v to see when and which objects, archives and dsos are loaded.
ld32: INFO 152: Output file removed because of error.

Any help or suggestions on debugging would be helpful! :D
:O2: - trajan - R5000 180Mhz - 256 MB RAM - 300GB HDD
:O2: - hadrian - R5000 180Mhz - 128 MB RAM - 147 GB HDD
:Octane2: - augustus - R12000 400Mhz - 512 MB RAM - 300GB HDD - VPro6

User avatar
necron2600
Posts: 279
Joined: Tue May 26, 2015 10:38 pm
Location: Boston

Re: Compiling Zephyr - At my wits end

Unread postby necron2600 » Wed Nov 15, 2017 7:28 pm

From my experience.. some library has _et_list as a symbol.

I often simply run something like in /usr/lib32/
for i in *.so
do
echo $i
nm -o $i >> /tmp/list
done

I ran that in /usr/lib32
[47] |000102640530|000000000004|OBJT |GLOB |DEFAULT |MIPS_DATA|_et_list
shows up under libcom_err.so

So during linking, you may have luck by adding in -lcom_err

Also, note that libss.a is a static library?? Do you really need that?

User avatar
praetor242
Posts: 265
Joined: Thu Feb 23, 2017 5:41 pm
Location: Arlington, TX

Re: Compiling Zephyr - At my wits end

Unread postby praetor242 » Wed Nov 15, 2017 7:40 pm

So do I just put -lcom_err in my LDFLAGS?

I wasn't really thrilled about it compiling static libraries, but that's how the makefile does it, and makefiles are still a little bit foreign to me.
:O2: - trajan - R5000 180Mhz - 256 MB RAM - 300GB HDD
:O2: - hadrian - R5000 180Mhz - 128 MB RAM - 147 GB HDD
:Octane2: - augustus - R12000 400Mhz - 512 MB RAM - 300GB HDD - VPro6

User avatar
necron2600
Posts: 279
Joined: Tue May 26, 2015 10:38 pm
Location: Boston

Re: Compiling Zephyr - At my wits end

Unread postby necron2600 » Fri Nov 17, 2017 10:08 am

Possibly (regarding -lcom_err ) :)

Another debugging tip:
There should be more output before the ld32: WARNING 84 error.. typically 'gcc or cc' commands and such.
You could copy/paste that entire long gcc/cc line before the error and re-run manually (assuming you are in the correct directory in the source) and play with adding -lcom_err and -L ,etc.. arguments. If it works, you could formalize it in LDFLAGS.

User avatar
praetor242
Posts: 265
Joined: Thu Feb 23, 2017 5:41 pm
Location: Arlington, TX

Re: Compiling Zephyr - At my wits end

Unread postby praetor242 » Fri Nov 17, 2017 4:38 pm

YAY! :D I got it! Turns out it was was trying to link to the wrong libcom_err.so . It was trying to link to the one supplied by libss2 which does not work for some reason. Being able to step through what ld and gcc were doing helped.

Thanks for all your help, necron2000 <3
:O2: - trajan - R5000 180Mhz - 256 MB RAM - 300GB HDD
:O2: - hadrian - R5000 180Mhz - 128 MB RAM - 147 GB HDD
:Octane2: - augustus - R12000 400Mhz - 512 MB RAM - 300GB HDD - VPro6

User avatar
jimmer
Donor
Donor
Posts: 483
Joined: Tue Oct 12, 2004 3:54 pm
Location: London, Ingerlund

Re: Compiling Zephyr - At my wits end

Unread postby jimmer » Fri Nov 17, 2017 6:28 pm

Does this mean you'll be getting to class on time from now on? :)

User avatar
praetor242
Posts: 265
Joined: Thu Feb 23, 2017 5:41 pm
Location: Arlington, TX

Re: Compiling Zephyr - At my wits end

Unread postby praetor242 » Fri Nov 17, 2017 8:05 pm

jimmer wrote:Does this mean you'll be getting to class on time from now on? :)



AHAHAHAHAHAHA!!!!!

For being an MIT project, it's not the most intuitive thing to get going and it's really only portable-ish. I had to modify a few things to get it to see some defines set in IRIX's standard library.
:O2: - trajan - R5000 180Mhz - 256 MB RAM - 300GB HDD
:O2: - hadrian - R5000 180Mhz - 128 MB RAM - 147 GB HDD
:Octane2: - augustus - R12000 400Mhz - 512 MB RAM - 300GB HDD - VPro6


Return to “SGI: Development”

Who is online

Users browsing this forum: No registered users and 2 guests