[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Can't build current hurd CVS tree -- error from skbuff.c
From: |
David Walter |
Subject: |
Re: Can't build current hurd CVS tree -- error from skbuff.c |
Date: |
Fri, 13 Sep 2002 11:22:17 -0400 |
User-agent: |
Gnus/5.090007 (Oort Gnus v0.07) XEmacs/21.4 (Honest Recruiter, hurd-i386-debian) |
Clemmitt Sigler <siglercm@jrt.me.vt.edu> writes:
> Hi,
>
> I pulled the hurd CVS tree today and tried to build it in order to
> get the fix to isofs/inode.c made last month by Marcus. But my
> build failed. I tried it on a hurd 0.3/libc0.3 (upgraded) system, and
> also on a system installed from the J1 CDs. Just last Thursday I
> successfully built a CVS pull of hurd on a J1 CD system (but had a few
> self-inflicted misadventures requiring me to start from scratch :^).
>
> The error seems strange to me. It dies on line 67 of
> pfinet/linux-src/net/core/skbuff.c. It seems it doesn't like
> SPIN_LOCK_UNLOCKED as it says the initializer isn't constant.
> All the code containing SPIN_LOCK_UNLOCKED hasn't been touched
> in 2 years, so I guess the breakage comes from elsewhere. The
> exact error message is just one line:
>
> ---------------------------------------------------------------------
> linux-src/net/core/skbuff.c:67: initializer element is not constant
> ---------------------------------------------------------------------
I encountered this problem. It appears that there were no changes in
the pfinet sources that I can see, but -std=gnu99 has been added to
the CFLAGS options.
If you compile this one file without this option it builds fine.
I have assumed here that the -std option only affect the c
preprocessor and not binary format, I am running with the current
binary from that build.
Unfortunately I don't know the correct fix for the empty structure
definition used for SPIN_LOCK_UNLOCKED. This is surely not the right
way to build.
Please correct me if I'm wrong, on this.
--
/^\
\ / ASCII RIBBON CAMPAIGN
X AGAINST HTML MAIL
/ \