Re: [PATCH] Re: linux-2.6.19-rc5-g088406bc build #120 failed

From: Jeff Garzik
Date: Mon Nov 13 2006 - 10:15:56 EST


Paul Fulghum wrote:
Jeff Garzik wrote:
Toralf Förster wrote:

Hello,

the build with the attached .config failed, make ends with:
... UPD include/linux/compile.h
CC init/version.o
LD init/built-in.o
LD .tmp_vmlinux1
drivers/built-in.o: In function `hdlcdev_open':
synclink.c:(.text+0x650d5): undefined reference to `hdlc_open'
synclink.c:(.text+0x6510d): undefined reference to `hdlc_open'
...
synclink_cs.c:(.text+0x7aece): undefined reference to `hdlc_ioctl'
drivers/built-in.o: In function `hdlcdev_init':
synclink_cs.c:(.text+0x7b336): undefined reference to `alloc_hdlcdev'
drivers/built-in.o: In function `hdlcdev_exit':
synclink_cs.c:(.text+0x7b434): undefined reference to `unregister_hdlc_device'
make: *** [.tmp_vmlinux1] Error 1


Does this patch work for you?

Jeff

No, this patch is not acceptable.

This has been beaten to death in previous threads.
The problem is a mismatch in your kernel config between
generic hdlc (M) and synclink (Y).

synclink drivers can *optionally* support generic hdlc.
You *must* be able to build synclink driver without generic hdlc.
Because of this you *can't* just put in the generic hdlc dependency.

Several alternative patches were posted (3 or 4 months) ago.
No particular patch won the approval of all kernel developers,
so nothing was done.


The existing build breakage is unacceptable too. It's this bogosity

#ifdef CONFIG_HDLC_MODULE
#define CONFIG_HDLC 1
#endif

which directly causes the build breakage.

Unless someone fixes it The Right Way, my patch seems like the best we can do. People who complain "you must be able to build without hdlc" should step up and make it so. Currently, it is not so.

Jeff


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/