On Sat, Sep 09, 2000 at 03:03:27PM +0900, NIIBE Yutaka wrote:
> - }
> + if (optname == IP_PKTINFO || optname == IP_RECVTTL
> + || optname == IP_RECVTOS || optname == IP_RECVOPTS
> + || optname == IP_RETOPTS || optname == IP_TOS
> + || optname == IP_TTL || optname ==
David S. Miller wrote:
> Could you send me a patch which fixes the problem in this way?
Sure. Here is the one. The occurrence of the IP_XXX corresponds the
one in switch/case.
Thank you for your time,
--- linux-2.4.0-test8-pre6/net/ipv4/ip_sockglue.c Fri Aug 11 05:01:26 2000
+++ linux/
From: NIIBE Yutaka <[EMAIL PROTECTED]>
Date: Sat, 09 Sep 2000 12:18:28 +0900
I'd like to explain my point clearly. My point is that accessing
with get_user as int is questionable. In my case, it's string. I
don't think all the string argment to the kernel should be aligned.
Tha
Hi David,
I'd like to explain my point clearly. My point is that accessing with
get_user as int is questionable. In my case, it's string. I don't
think all the string argment to the kernel should be aligned.
David S. Miller wrote:
> Why not make sure in the user tools that the argument is p
From: NIIBE Yutaka <[EMAIL PROTECTED]>
Date:Sat, 09 Sep 2000 11:25:28 +0900
Here's a patch, avoiding useless access.
This seems like a large ugly hack.
Why not make sure in the user tools that the argument is properly
aligned to 4 bytes?
IP-chains works fine on Alpha and Sparc
With ipchains, we have alignment problem. H. Kambara
<[EMAIL PROTECTED]> found that it core dumps on SuperH machine.
The cause of this problem is get_user accesses wrongly in
ip_setsockopt.
Here's a patch, avoiding useless access.
diff -ruN linux-2.4.0-test8-pre6/net/ipv4/ip_sockglue.c kernel/
6 matches
Mail list logo
