thr3ads.net - netfilter buglog - [Bug 1347] New: ebtables-nft: regression in -o option [Jul 2019]

If this information is useful, please help other people find it:
Share via:

bugzilla-daemon at netfilter.org

2019-Jul-02 08:50 UTC

[Bug 1347] New: ebtables-nft: regression in -o option

https://bugzilla.netfilter.org/show_bug.cgi?id=1347

            Bug ID: 1347
           Summary: ebtables-nft: regression in -o option
           Product: nftables
           Version: unspecified
          Hardware: x86_64
                OS: Debian GNU/Linux
            Status: NEW
          Severity: normal
          Priority: P5
         Component: iptables over nftable
          Assignee: pablo at netfilter.org
          Reporter: arturo at debian.org

This ruleset:

# cat ebtables-fwd-no-o-options-allowed.rules
*filter
:PVEFW-FORWARD ACCEPT
:PVEFW-FWBR-OUT ACCEPT
-A PVEFW-FORWARD -p IPv4 -j ACCEPT
-A PVEFW-FORWARD -p IPv6 -j ACCEPT
-A PVEFW-FORWARD -o fwln+ -j PVEFW-FWBR-OUT
-A FORWARD -j PVEFW-FORWARD

works in ebtables-legacy but doesn't in ebtables-nft.

# cat ebtables-fwd-no-o-options-allowed.rules |  ebtables-legacy-restore
# cat ebtables-fwd-no-o-options-allowed.rules |  ebtables-nft-restore
Use -o only in OUTPUT, FORWARD and POSTROUTING chains.

More context in the original Debian bug: https://bugs.debian.org/929976

-- 
You are receiving this mail because:
You are watching all bug changes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.netfilter.org/pipermail/netfilter-buglog/attachments/20190702/0d54b257/attachment.html>

bugzilla-daemon at netfilter.org

2019-Jul-02 10:58 UTC

head link

[Bug 1347] ebtables-nft: regression in -o option

https://bugzilla.netfilter.org/show_bug.cgi?id=1347

Florian Westphal <fw at strlen.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |fw at strlen.de

--- Comment #1 from Florian Westphal <fw at strlen.de> ---
(In reply to Arturo Borrero Gonzalez from comment #0)> This ruleset:
> 
> # cat ebtables-fwd-no-o-options-allowed.rules
> *filter
> :PVEFW-FORWARD ACCEPT
> :PVEFW-FWBR-OUT ACCEPT
> -A PVEFW-FORWARD -p IPv4 -j ACCEPT
> -A PVEFW-FORWARD -p IPv6 -j ACCEPT
> -A PVEFW-FORWARD -o fwln+ -j PVEFW-FWBR-OUT
> -A FORWARD -j PVEFW-FORWARD
> 
> works in ebtables-legacy but doesn't in ebtables-nft.
Fix:

diff --git a/iptables/xtables-eb.c b/iptables/xtables-eb.c
--- a/iptables/xtables-eb.c
+++ b/iptables/xtables-eb.c
@@ -197,7 +197,8 @@ int ebt_get_current_chain(const char *chain)
        else if (strcmp(chain, "POSTROUTING") == 0)
                return NF_BR_POST_ROUTING;

-       return -1;
+       /* placeholder for user defined chain */
+       return NF_BR_NUMHOOKS;
 }

Let me know and I'll push it to iptables.git.

-- 
You are receiving this mail because:
You are watching all bug changes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.netfilter.org/pipermail/netfilter-buglog/attachments/20190702/70248d55/attachment.html>

bugzilla-daemon at netfilter.org

2019-Jul-02 11:13 UTC

head link

[Bug 1347] ebtables-nft: regression in -o option

https://bugzilla.netfilter.org/show_bug.cgi?id=1347

--- Comment #2 from Arturo Borrero Gonzalez <arturo at debian.org> ---
(In reply to Florian Westphal from comment #1)> (In reply to Arturo Borrero Gonzalez from comment #0)
> > This ruleset:
> > 
> > # cat ebtables-fwd-no-o-options-allowed.rules
> > *filter
> > :PVEFW-FORWARD ACCEPT
> > :PVEFW-FWBR-OUT ACCEPT
> > -A PVEFW-FORWARD -p IPv4 -j ACCEPT
> > -A PVEFW-FORWARD -p IPv6 -j ACCEPT
> > -A PVEFW-FORWARD -o fwln+ -j PVEFW-FWBR-OUT
> > -A FORWARD -j PVEFW-FORWARD
> > 
> > works in ebtables-legacy but doesn't in ebtables-nft.
> 
> Fix:
> 
> diff --git a/iptables/xtables-eb.c b/iptables/xtables-eb.c
> --- a/iptables/xtables-eb.c
> +++ b/iptables/xtables-eb.c
> @@ -197,7 +197,8 @@ int ebt_get_current_chain(const char *chain)
>         else if (strcmp(chain, "POSTROUTING") == 0)
>                 return NF_BR_POST_ROUTING;
>  
> -       return -1;
> +       /* placeholder for user defined chain */
> +       return NF_BR_NUMHOOKS;
>  }
> 
> Let me know and I'll push it to iptables.git.
Something is wrong with the change. ebtables-nft-restore seems to be in some
busy loop consuming 100% CPU. If called with strace, the behavior changes:

[..]
stat("/usr/local/lib/xtables/libebt_standard.so", 0x7ffc0d6b1ca0) = -1
ENOENT
(No such file or directory)
stat("/usr/local/lib/xtables/libxt_standard.so",
{st_mode=S_IFREG|0755,
st_size=25136, ...}) = 0
openat(AT_FDCWD, "/usr/local/lib/xtables/libxt_standard.so",
O_RDONLY|O_CLOEXEC) = 4
read(4,
"\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0`\20\0\0\0\0\0\0"...,
832) = 832
fstat(4, {st_mode=S_IFREG|0755, st_size=25136, ...}) = 0
mmap(NULL, 16688, PROT_READ, MAP_PRIVATE|MAP_DENYWRITE, 4, 0) = 0x7efc36c5f000
mmap(0x7efc36c60000, 4096, PROT_READ|PROT_EXEC,
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 4, 0x1000) = 0x7efc36c60000
mmap(0x7efc36c61000, 4096, PROT_READ, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 4,
0x2000) = 0x7efc36c61000
mmap(0x7efc36c62000, 8192, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 4, 0x2000) = 0x7efc36c62000
close(4)                                = 0
mprotect(0x7efc36c62000, 4096, PROT_READ) = 0
fstat(0, {st_mode=S_IFCHR|0620, st_rdev=makedev(0x88, 0x8), ...}) = 0
read(0, <--- waits here

-- 
You are receiving this mail because:
You are watching all bug changes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.netfilter.org/pipermail/netfilter-buglog/attachments/20190702/55ac63bd/attachment.html>

bugzilla-daemon at netfilter.org

2019-Jul-02 11:27 UTC

head link

[Bug 1347] ebtables-nft: regression in -o option

https://bugzilla.netfilter.org/show_bug.cgi?id=1347

--- Comment #3 from Florian Westphal <fw at strlen.de> ---
(In reply to Arturo Borrero Gonzalez from comment #2)> (In reply to Florian Westphal from comment #1)
> > (In reply to Arturo Borrero Gonzalez from comment #0)
> > > This ruleset:
> > > 
> > > # cat ebtables-fwd-no-o-options-allowed.rules
> > > *filter
> > > :PVEFW-FORWARD ACCEPT
> > > :PVEFW-FWBR-OUT ACCEPT
> > > -A PVEFW-FORWARD -p IPv4 -j ACCEPT
> > > -A PVEFW-FORWARD -p IPv6 -j ACCEPT
> > > -A PVEFW-FORWARD -o fwln+ -j PVEFW-FWBR-OUT
> > > -A FORWARD -j PVEFW-FORWARD
> > > 
> > > works in ebtables-legacy but doesn't in ebtables-nft.
> > 
> > Fix:
> > 
> > diff --git a/iptables/xtables-eb.c b/iptables/xtables-eb.c
> > --- a/iptables/xtables-eb.c
> > +++ b/iptables/xtables-eb.c
> > @@ -197,7 +197,8 @@ int ebt_get_current_chain(const char *chain)
> >         else if (strcmp(chain, "POSTROUTING") == 0)
> >                 return NF_BR_POST_ROUTING;
> >  
> > -       return -1;
> > +       /* placeholder for user defined chain */
> > +       return NF_BR_NUMHOOKS;
> >  }
> > 
> > Let me know and I'll push it to iptables.git.
> 
> Something is wrong with the change. ebtables-nft-restore seems to be in
some
> busy loop consuming 100% CPU. If called with strace, the behavior changes:
> 
> [..]
> stat("/usr/local/lib/xtables/libebt_standard.so", 0x7ffc0d6b1ca0)
= -1
> ENOENT (No such file or directory)
> stat("/usr/local/lib/xtables/libxt_standard.so",
{st_mode=S_IFREG|0755,
> st_size=25136, ...}) = 0
> openat(AT_FDCWD, "/usr/local/lib/xtables/libxt_standard.so",
> O_RDONLY|O_CLOEXEC) = 4
> read(4,
"\177ELF\2\1\1\0\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0`\20\0\0\0\0\0\0"...,
> 832) = 832
> fstat(4, {st_mode=S_IFREG|0755, st_size=25136, ...}) = 0
> mmap(NULL, 16688, PROT_READ, MAP_PRIVATE|MAP_DENYWRITE, 4, 0) >
0x7efc36c5f000
> mmap(0x7efc36c60000, 4096, PROT_READ|PROT_EXEC,
> MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 4, 0x1000) = 0x7efc36c60000
> mmap(0x7efc36c61000, 4096, PROT_READ, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE,
> 4, 0x2000) = 0x7efc36c61000
> mmap(0x7efc36c62000, 8192, PROT_READ|PROT_WRITE,
> MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 4, 0x2000) = 0x7efc36c62000
> close(4)                                = 0
> mprotect(0x7efc36c62000, 4096, PROT_READ) = 0
> fstat(0, {st_mode=S_IFCHR|0620, st_rdev=makedev(0x88, 0x8), ...}) = 0
> read(0, <--- waits here
How can I reproduce this?

It works for me, iptables-test.py --nftables passes too.

export XTABLES_LIBDIR=$(pwd)/extensions
unshare -n sh -c "iptables/xtables-nft-multi ebtables-restore < t;
iptables/xtables-nft-multi ebtables-save ; nft list ruleset"
# Generated by ebtables-save v1.8.3 on Tue Jul  2 13:21:56 2019
*filter
:INPUT ACCEPT
:FORWARD ACCEPT
:OUTPUT ACCEPT
:PVEFW-FORWARD ACCEPT
:PVEFW-FWBR-OUT ACCEPT
-A FORWARD -j PVEFW-FORWARD
-A PVEFW-FORWARD -p IPv4 -j ACCEPT
-A PVEFW-FORWARD -p IPv6 -j ACCEPT
-A PVEFW-FORWARD -o fwln+ -j PVEFW-FWBR-OUT

table bridge filter {
        chain INPUT {
                type filter hook input priority filter; policy accept;
        }

        chain FORWARD {
                type filter hook forward priority filter; policy accept;
                counter packets 0 bytes 0 jump PVEFW-FORWARD
        }

        chain OUTPUT {
                type filter hook output priority filter; policy accept;
        }

        chain PVEFW-FORWARD {
                ether type ip counter packets 0 bytes 0 accept
                ether type ip6 counter packets 0 bytes 0 accept
                oifname "fwln*" counter packets 0 bytes 0 jump
PVEFW-FWBR-OUT
                counter packets 0 bytes 0 accept
        }

        chain PVEFW-FWBR-OUT {
                counter packets 0 bytes 0 accept
        }
}

-- 
You are receiving this mail because:
You are watching all bug changes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.netfilter.org/pipermail/netfilter-buglog/attachments/20190702/2db016d5/attachment.html>

bugzilla-daemon at netfilter.org

2020-Nov-17 12:03 UTC

head link

[Bug 1347] ebtables-nft: regression in -o option

https://bugzilla.netfilter.org/show_bug.cgi?id=1347

Phil Sutter <phil at nwl.cc> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |phil at nwl.cc

--- Comment #4 from Phil Sutter <phil at nwl.cc> ---
Arturo, is this still relevant?

-- 
You are receiving this mail because:
You are watching all bug changes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.netfilter.org/pipermail/netfilter-buglog/attachments/20201117/e5661633/attachment.html>

bugzilla-daemon at netfilter.org

2020-Nov-17 12:05 UTC

head link

[Bug 1347] ebtables-nft: regression in -o option

https://bugzilla.netfilter.org/show_bug.cgi?id=1347

Arturo Borrero Gonzalez <arturo at netfilter.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #5 from Arturo Borrero Gonzalez <arturo at netfilter.org> ---
(In reply to Phil Sutter from comment #4)> Arturo, is this still relevant?
It is fixed apparently! thanks!

-- 
You are receiving this mail because:
You are watching all bug changes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.netfilter.org/pipermail/netfilter-buglog/attachments/20201117/d27f0a80/attachment.html>

Seemingly Similar Threads

Search for more possibly parallel threads

netfilter buglog - Jul 2019 - [Bug 1347] New: ebtables-nft: regression in -o option

[Bug 1347] New: ebtables-nft: regression in -o option

[Bug 1347] ebtables-nft: regression in -o option

[Bug 1347] ebtables-nft: regression in -o option

[Bug 1347] ebtables-nft: regression in -o option

[Bug 1347] ebtables-nft: regression in -o option

[Bug 1347] ebtables-nft: regression in -o option

Seemingly Similar Threads