LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: PATCH: modules_install and modules_uninstall target in Makefile

To: lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: PATCH: modules_install and modules_uninstall target in Makefile
From: Horms <horms@xxxxxxxxxxxx>
Date: Sat, 8 Jun 2002 12:14:08 +0900
On Sat, Jun 08, 2002 at 02:25:57AM +0200, Roberto Nibali wrote:
> Yo Horms,
> 
> > CC=gcc
> > CFLAGS= -D__KERNEL__ -DMODULE -DEXPORT_SYMTAB -DMODVERSIONS \
> >-    $(SMPFLAGS) $(DEBUGFLAGS) -O2 -Wall \
> >-    -Wstrict-prototypes -I/usr/src/linux/include \
> >+    $(SMPFLAGS) $(DEBUGFLAGS) -O2 -Wall -Wstrict-prototypes \
> >+    -I/home/horms/lang/c/kernel/linux-2.4.18/include \
> 
> :) I'm sure you don't want this.

Well I did, but I'm pretty sure no-one else does :)
That said, IMHO the Makefile should provide some way
for the user to pass in aditional include paths.

> > install:    ip_vs.o
> >     -rmmod ip_vs
> >     insmod ip_vs.o
> >     lsmod
> >+
> >+modules_install:
> >+    if [ ! -d "$(MODLIB)" ]; then mkdir -p "$(MODLIB)"; fi
> >+    install -m 600 -c $(CORE) $(SCHEDULERS) "$(MODLIB)"
> >+    depmod -a
> >+
> >+modules_uninstall:
> >+    (cd "$(MODLIB)"; rm $(CORE) $(SCHEDULERS))
> 
> s/rm/rm -f/, just to make sure and for the eye ;)

Good call.

I was also thinking that modules_install should depend on all

New patch attached.

-- 
Horms
        

Attachment: ipvs-1.0.3.ipvs.Makefile.2.patch
Description: Text document

<Prev in Thread] Current Thread [Next in Thread>