Qmail - Compiling Fastforward errors

This is Interesting: Free IT Magazines  
Home > Archive > Qmail > March 2006 > Compiling Fastforward errors





You are viewing an archived Text-only version of the thread. To view this thread in it's original format and/or if you want to reply to this thread please [click here]

Author Compiling Fastforward errors
Andrew Kenna

2006-03-21, 9:01 am

Anyone know a different link that I can download fastforward from... I've
just installed a fresh server and compiled up qmail and all its associated
packages fine as per the Life Without Qmail document.

I've downloaded it from the following website
http://cr.yp.to/fastforward.html

When I got to compiling Fastforward it comes up with the following error

[root@calaghan fastforward-0.51]# make setup check
../load auto-str substdio.a error.a str.a
/usr/bin/ld: errno: TLS definition in /lib/libc.so.6 section .tbss
mismatches non-TLS reference in substdio.a(substdo.o)
/lib/libc.so.6: could not read symbols: Bad value
collect2: ld returned 1 exit status
make: *** [auto-str] Error 1

I am running Redhat Fedora Core 4
GCC Version 4.0.2
GNU LD Version 2.15.94.0.2.2 20041220

Does anyone have any suggestions as to why this error could be coming up ? I
have not patched Fastforward with anything, I have applied the patches to
qmail and ucsp as required in the LWQ Document.

--
Regards

Andrew Kenna
Ajk Communications
ABN: 29 831 556 566
email: akenna@westnet.com.au
web: www.ajkcommunications.com.au


AK

2006-03-21, 9:01 am

Andrew Kenna wrote:

> Anyone know a different link that I can download fastforward from... I've
> just installed a fresh server and compiled up qmail and all its associated
> packages fine as per the Life Without Qmail document.
>
> I've downloaded it from the following website
> http://cr.yp.to/fastforward.html
>
> When I got to compiling Fastforward it comes up with the following error
>
> [root@calaghan fastforward-0.51]# make setup check
> ./load auto-str substdio.a error.a str.a
> /usr/bin/ld: errno: TLS definition in /lib/libc.so.6 section .tbss
> mismatches non-TLS reference in substdio.a(substdo.o)
> /lib/libc.so.6: could not read symbols: Bad value
> collect2: ld returned 1 exit status
> make: *** [auto-str] Error 1
>
> I am running Redhat Fedora Core 4
> GCC Version 4.0.2
> GNU LD Version 2.15.94.0.2.2 20041220
>
> Does anyone have any suggestions as to why this error could be coming up ? I
> have not patched Fastforward with anything, I have applied the patches to
> qmail and ucsp as required in the LWQ Document.
>
> --
> Regards
>
> Andrew Kenna
> Ajk Communications
> ABN: 29 831 556 566
> email: akenna@westnet.com.au
> web: www.ajkcommunications.com.au
>
>


I believe there was a similar question posted a while back,

The error is due to the explicit definition of a system variable errno.
Find the header file that defines extern int errno; and replace the line
with #include <errno.h>.

This should solve your problem.

AK
Andrew Kenna

2006-03-21, 9:01 am

yes the exact same question was posted by me, and i thought i would solve it
by loading up a fresh server with nothing else on it except qmail and the
operating system.

Nothing made any differences..

Would this problem be within fastforward or qmail or the actual operating
system itself ?

Andrew

"AK" <aktrader2@excite.com> wrote in message
news:KeSdnf9qy9euHYLZnZ2dnUVZ_tidnZ2d@co
mcast.com...
> Andrew Kenna wrote:
>
I've[vbcol=seagreen]
associated[vbcol=seagreen]
? I[vbcol=seagreen]
to[vbcol=seagreen]
>
> I believe there was a similar question posted a while back,
>
> The error is due to the explicit definition of a system variable errno.
> Find the header file that defines extern int errno; and replace the line
> with #include <errno.h>.
>
> This should solve your problem.
>
> AK



Andrew Kenna

2006-03-21, 9:01 am

Found the problem, It appears that its in the error.h file within the
fastforward package.

thanks for your assistance AK

Does everyone else have this problem when compiling fastforward or is it
something unique with fedora core 4

andrew

"AK" <aktrader2@excite.com> wrote in message
news:KeSdnf9qy9euHYLZnZ2dnUVZ_tidnZ2d@co
mcast.com...
> Andrew Kenna wrote:
>
I've[vbcol=seagreen]
associated[vbcol=seagreen]
? I[vbcol=seagreen]
to[vbcol=seagreen]
>
> I believe there was a similar question posted a while back,
>
> The error is due to the explicit definition of a system variable errno.
> Find the header file that defines extern int errno; and replace the line
> with #include <errno.h>.
>
> This should solve your problem.
>
> AK



AK

2006-03-21, 9:01 am

Andrew Kenna wrote:

> Found the problem, It appears that its in the error.h file within the
> fastforward package.
>
> thanks for your assistance AK
>
> Does everyone else have this problem when compiling fastforward or is it
> something unique with fedora core 4
>
> andrew
>
> "AK" <aktrader2@excite.com> wrote in message
> news:KeSdnf9qy9euHYLZnZ2dnUVZ_tidnZ2d@co
mcast.com...
>
>
> I've
>
>
> associated
>
>
> ? I
>
>
> to
>
>
>
>



With the newer versions of gcc >3.x the compilers rejects or generates
errors when a "system" variable is being explicitly redefined in the
software being compiled.

I believe all DJB's software packages have this issue where errno is
explicitly defined as extern int errno;. The remedy is to replace this
line within any header file with the #include <errno.h>.

The initial 3.x version of gcc would reject this redefinition. some
subsequent gcc versions merely generated warnings that a redifinition
has taken place.
If you have successfully compiled a package with the explicit definition
of errno and then you upgrade the gcc package, the application will
likely start crashing due to the errno issue (saw it happen).


AK
Sponsored Links






Free braindumps | Software forum | Database administration forum

Copyright 2003 - 2009 webservertalk.com