linux-decnet-user — Discussion for users of DECnet on Linux

[Linux-decnet-user] [-mm patch] net/decnet/: cleanups

[Adrian B. <bu...@st...>]

On Sun, Aug 13, 2006 at 01:24:54AM -0700, Andrew Morton wrote:
>...
> Changes since 2.6.18-rc3-mm2:
>...
>  git-net.patch
>...
>  git trees
>...

This patch contains the following cleanups:
- make the following needlessly global functions static:
  - dn_fib.c: dn_fib_sync_down()
  - dn_fib.c: dn_fib_sync_up()
  - dn_rules.c: dn_fib_rule_action()
- remove the following unneeded prototype:
  - dn_fib.c: dn_cache_dump()

Signed-off-by: Adrian Bunk <bu...@st...>

---

 include/net/dn_fib.h  |    3 ---
 net/decnet/dn_fib.c   |    9 +++++----
 net/decnet/dn_rules.c |    4 ++--
 3 files changed, 7 insertions(+), 9 deletions(-)

--- linux-2.6.18-rc4-mm1/include/net/dn_fib.h.old	2006-08-13 20:18:58.000000000 +0200
+++ linux-2.6.18-rc4-mm1/include/net/dn_fib.h	2006-08-13 20:19:08.000000000 +0200
@@ -131,9 +131,6 @@
 extern void dn_fib_flush(void);
 extern void dn_fib_select_multipath(const struct flowi *fl,
 					struct dn_fib_res *res);
-extern int dn_fib_sync_down(__le16 local, struct net_device *dev,
-				int force);
-extern int dn_fib_sync_up(struct net_device *dev);
 
 /*
  * dn_tables.c
--- linux-2.6.18-rc4-mm1/net/decnet/dn_fib.c.old	2006-08-13 20:20:05.000000000 +0200
+++ linux-2.6.18-rc4-mm1/net/decnet/dn_fib.c	2006-08-13 20:22:40.000000000 +0200
@@ -55,8 +55,6 @@
 
 #define endfor_nexthops(fi) }
 
-extern int dn_cache_dump(struct sk_buff *skb, struct netlink_callback *cb);
-
 static DEFINE_SPINLOCK(dn_fib_multipath_lock);
 static struct dn_fib_info *dn_fib_info_list;
 static DEFINE_SPINLOCK(dn_fib_info_lock);
@@ -80,6 +78,9 @@
 	[RTN_XRESOLVE] =    { .error = -EINVAL, .scope = RT_SCOPE_NOWHERE },
 };
 
+static int dn_fib_sync_down(__le16 local, struct net_device *dev, int force);
+static int dn_fib_sync_up(struct net_device *dev);
+
 void dn_fib_free_info(struct dn_fib_info *fi)
 {
 	if (fi->fib_dead == 0) {
@@ -651,7 +652,7 @@
 	return NOTIFY_DONE;
 }
 
-int dn_fib_sync_down(__le16 local, struct net_device *dev, int force)
+static int dn_fib_sync_down(__le16 local, struct net_device *dev, int force)
 {
         int ret = 0;
         int scope = RT_SCOPE_NOWHERE;
@@ -695,7 +696,7 @@
 }
 
 
-int dn_fib_sync_up(struct net_device *dev)
+static int dn_fib_sync_up(struct net_device *dev)
 {
         int ret = 0;
 
--- linux-2.6.18-rc4-mm1/net/decnet/dn_rules.c.old	2006-08-13 20:22:45.000000000 +0200
+++ linux-2.6.18-rc4-mm1/net/decnet/dn_rules.c	2006-08-13 20:23:02.000000000 +0200
@@ -75,8 +75,8 @@
 	return err;
 }
 
-int dn_fib_rule_action(struct fib_rule *rule, struct flowi *flp, int flags,
-		       struct fib_lookup_arg *arg)
+static int dn_fib_rule_action(struct fib_rule *rule, struct flowi *flp,
+			      int flags, struct fib_lookup_arg *arg)
 {
 	int err = -EAGAIN;
 	struct dn_fib_table *tbl;

Re: [Linux-decnet-user] New Xorg rpms uploaded

[Patrick C. <pa...@ty...>]

Patrick Caulfield wrote:
> I've made a small fix to the DECnet-enabled Xorg rpms on the web site. It
> fixes a problem where busy applications could quit because the kernel returned
> -1/EINVAL from write().
> 
> http://linux-decnet.sourceforge.net/Xservers/

Those didn't fix the problem, but the new ones now up seem to. Sorry about that.

-- 

patrick

[Linux-decnet-user] latd 1.25 released

[Patrick C. <pa...@ty...>]

This is just some minor fixes.

  * llogin stops terminal driver generating CRs. Should help in editors.
  * Slightly improve a couple of error messages.

and also to get the gcc 4.1 fixes out into the open.
-- 

patrick

[Linux-decnet-user] New Xorg rpms uploaded

[Patrick C. <pa...@ty...>]

I've made a small fix to the DECnet-enabled Xorg rpms on the web site. It
fixes a problem where busy applications could quit because the kernel returned
-1/EINVAL from write().

http://linux-decnet.sourceforge.net/Xservers/

Thanks to Adam Pigg for spotting the problem and supplying strace evidence - I
hope this fixes it for you!

-- 

patrick

Re: [Linux-decnet-user] Build error in dnprogs 2.37

[Patrick C. <pa...@ty...>]

ada...@br... wrote:
> Lo
> 
> This may or may not be a real build error on some platforms but....
> libvaxdata fails to build for me on a rhel3 system.  The error is
> libvaxdata.a not such file.  I tracked the problem to 'ar'.  The man page
> states that the -c option (in makefile.linux) supresses a warning about
> creating an archive, even though an archive is always created.  For some
> reason, in my case, the -c option causes the archive to not be created, and
> removing this option from the command results in a successful build.
> 
> This is a stock rhel3 system.
> 
> Feel free to forward to the list.

Thanks, I've removed the "-c" in CVS.

odd bug though!

-- 

patrick