* [Qemu-devel] hostfwd_add and -netdev user
@ 2011-11-16 10:07 Markus Armbruster
2011-11-18 12:37 ` Jan Kiszka
0 siblings, 1 reply; 4+ messages in thread
From: Markus Armbruster @ 2011-11-16 10:07 UTC (permalink / raw)
To: qemu-devel; +Cc: jan.kiszka, Mark McLoughlin
Maybe I'm missing something, but it looks like hostfwd_add doesn't fully
work with -netdev.
Command definition from hmp-commands.hx:
{
.name = "hostfwd_add",
.args_type = "arg1:s,arg2:s?,arg3:s?",
.params = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
.help = "redirect TCP or UDP connections from host to guest (requires -net user)",
.mhandler.cmd = net_slirp_hostfwd_add,
},
Note: params says command takes either 1 or 3 arguments. args_type
can't express that, so it asks for 1-3.
Command handler:
void net_slirp_hostfwd_add(Monitor *mon, const QDict *qdict)
{
const char *redir_str;
SlirpState *s;
const char *arg1 = qdict_get_str(qdict, "arg1");
const char *arg2 = qdict_get_try_str(qdict, "arg2");
const char *arg3 = qdict_get_try_str(qdict, "arg3");
if (arg2) {
s = slirp_lookup(mon, arg1, arg2);
redir_str = arg3;
} else {
s = slirp_lookup(mon, NULL, NULL);
redir_str = arg1;
}
if (s) {
slirp_hostfwd(s, redir_str, 0);
}
}
If I understand the code correctly, the command has the following forms:
* hostfwd_add REDIR
This form uses the default stack (first member of slirp_stacks).
Seems to work okay with -netdev:
$ qemu-system-x86_64 --nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -netdev user,id=net.0 -device virtio-net-pci,netdev=net.0,id=nic.0
QEMU 0.15.92 monitor - type 'help' for more information
(qemu) hostfwd_add tcp::12345-:22
(qemu) info usernet
VLAN -1 (net.0):
Protocol[State] FD Source Address Port Dest. Address Port RecvQ SendQ
TCP[HOST_FORWARD] 12 * 12345 10.0.2.15 22 0 0
* hostfwd_add VLAN STACK REDIR
This form was added in commit f13b572c "slirp: Make hostfwd_add/remove
multi-instance-aware".
I don't like how it overloads the command; adding the optional
arguments at the end would have been cleaner. Water under the bridge.
Note that slirp_lookup() converts VLAN to an int (without error
checking, i.e. non-integers are silently interpreted as zero, integers
out of bounds are silently truncated).
Anyway, I can't make it work with -netdev:
(qemu) hostfwd_add -1 net.0 tcp::12345-:22
unknown VLAN -1
* Just for giggles: what if I give two arguments?
$ qemu-system-x86_64--nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -net user -device virtio-net-pci,id=nic.0,vlan=0QEMU 0.15.92 monitor - type 'help' for more information
(qemu) hostfwd_add 0 user.0
invalid host forwarding rule '(null)'
The (null) is from a printf-like function printing a null pointer with
conversion %s. Crash bug on some systems.
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [Qemu-devel] hostfwd_add and -netdev user
2011-11-16 10:07 [Qemu-devel] hostfwd_add and -netdev user Markus Armbruster
@ 2011-11-18 12:37 ` Jan Kiszka
2011-11-18 13:07 ` Markus Armbruster
0 siblings, 1 reply; 4+ messages in thread
From: Jan Kiszka @ 2011-11-18 12:37 UTC (permalink / raw)
To: Markus Armbruster; +Cc: Mark McLoughlin, qemu-devel
[-- Attachment #1: Type: text/plain, Size: 3334 bytes --]
On 2011-11-16 08:07, Markus Armbruster wrote:
> Maybe I'm missing something, but it looks like hostfwd_add doesn't fully
> work with -netdev.
>
> Command definition from hmp-commands.hx:
>
> {
> .name = "hostfwd_add",
> .args_type = "arg1:s,arg2:s?,arg3:s?",
> .params = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
> .help = "redirect TCP or UDP connections from host to guest (requires -net user)",
> .mhandler.cmd = net_slirp_hostfwd_add,
> },
>
> Note: params says command takes either 1 or 3 arguments. args_type
> can't express that, so it asks for 1-3.
>
> Command handler:
>
> void net_slirp_hostfwd_add(Monitor *mon, const QDict *qdict)
> {
> const char *redir_str;
> SlirpState *s;
> const char *arg1 = qdict_get_str(qdict, "arg1");
> const char *arg2 = qdict_get_try_str(qdict, "arg2");
> const char *arg3 = qdict_get_try_str(qdict, "arg3");
>
> if (arg2) {
> s = slirp_lookup(mon, arg1, arg2);
> redir_str = arg3;
> } else {
> s = slirp_lookup(mon, NULL, NULL);
> redir_str = arg1;
> }
> if (s) {
> slirp_hostfwd(s, redir_str, 0);
> }
>
> }
>
> If I understand the code correctly, the command has the following forms:
>
> * hostfwd_add REDIR
>
> This form uses the default stack (first member of slirp_stacks).
> Seems to work okay with -netdev:
>
> $ qemu-system-x86_64 --nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -netdev user,id=net.0 -device virtio-net-pci,netdev=net.0,id=nic.0
> QEMU 0.15.92 monitor - type 'help' for more information
> (qemu) hostfwd_add tcp::12345-:22
> (qemu) info usernet
> VLAN -1 (net.0):
> Protocol[State] FD Source Address Port Dest. Address Port RecvQ SendQ
> TCP[HOST_FORWARD] 12 * 12345 10.0.2.15 22 0 0
>
> * hostfwd_add VLAN STACK REDIR
>
> This form was added in commit f13b572c "slirp: Make hostfwd_add/remove
> multi-instance-aware".
>
> I don't like how it overloads the command; adding the optional
> arguments at the end would have been cleaner. Water under the bridge.
>
> Note that slirp_lookup() converts VLAN to an int (without error
> checking, i.e. non-integers are silently interpreted as zero, integers
> out of bounds are silently truncated).
>
> Anyway, I can't make it work with -netdev:
>
> (qemu) hostfwd_add -1 net.0 tcp::12345-:22
> unknown VLAN -1
>
> * Just for giggles: what if I give two arguments?
>
> $ qemu-system-x86_64--nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -net user -device virtio-net-pci,id=nic.0,vlan=0QEMU 0.15.92 monitor - type 'help' for more information
> (qemu) hostfwd_add 0 user.0
> invalid host forwarding rule '(null)'
>
> The (null) is from a printf-like function printing a null pointer with
> conversion %s. Crash bug on some systems.
>
Agreed, there is some need for improvements. I would not mind changing
the command syntax, in fact, if it helps and no major user is already
relying on it (can't imagine, specifically as netdev support is broken).
What about
hostfwd_add rule [name [vlan]]
?
Jan
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [Qemu-devel] hostfwd_add and -netdev user
2011-11-18 12:37 ` Jan Kiszka
@ 2011-11-18 13:07 ` Markus Armbruster
2011-11-18 13:13 ` Jan Kiszka
0 siblings, 1 reply; 4+ messages in thread
From: Markus Armbruster @ 2011-11-18 13:07 UTC (permalink / raw)
To: Jan Kiszka; +Cc: Mark McLoughlin, qemu-devel
Jan Kiszka <jan.kiszka@web.de> writes:
> On 2011-11-16 08:07, Markus Armbruster wrote:
>> Maybe I'm missing something, but it looks like hostfwd_add doesn't fully
>> work with -netdev.
>>
>> Command definition from hmp-commands.hx:
>>
>> {
>> .name = "hostfwd_add",
>> .args_type = "arg1:s,arg2:s?,arg3:s?",
>> .params = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
>> .help = "redirect TCP or UDP connections from host to guest (requires -net user)",
>> .mhandler.cmd = net_slirp_hostfwd_add,
>> },
>>
>> Note: params says command takes either 1 or 3 arguments. args_type
>> can't express that, so it asks for 1-3.
>>
>> Command handler:
>>
>> void net_slirp_hostfwd_add(Monitor *mon, const QDict *qdict)
>> {
>> const char *redir_str;
>> SlirpState *s;
>> const char *arg1 = qdict_get_str(qdict, "arg1");
>> const char *arg2 = qdict_get_try_str(qdict, "arg2");
>> const char *arg3 = qdict_get_try_str(qdict, "arg3");
>>
>> if (arg2) {
>> s = slirp_lookup(mon, arg1, arg2);
>> redir_str = arg3;
>> } else {
>> s = slirp_lookup(mon, NULL, NULL);
>> redir_str = arg1;
>> }
>> if (s) {
>> slirp_hostfwd(s, redir_str, 0);
>> }
>>
>> }
>>
>> If I understand the code correctly, the command has the following forms:
>>
>> * hostfwd_add REDIR
>>
>> This form uses the default stack (first member of slirp_stacks).
>> Seems to work okay with -netdev:
>>
>> $ qemu-system-x86_64 --nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -netdev user,id=net.0 -device virtio-net-pci,netdev=net.0,id=nic.0
>> QEMU 0.15.92 monitor - type 'help' for more information
>> (qemu) hostfwd_add tcp::12345-:22
>> (qemu) info usernet
>> VLAN -1 (net.0):
>> Protocol[State] FD Source Address Port Dest. Address Port RecvQ SendQ
>> TCP[HOST_FORWARD] 12 * 12345 10.0.2.15 22 0 0
>>
>> * hostfwd_add VLAN STACK REDIR
>>
>> This form was added in commit f13b572c "slirp: Make hostfwd_add/remove
>> multi-instance-aware".
>>
>> I don't like how it overloads the command; adding the optional
>> arguments at the end would have been cleaner. Water under the bridge.
>>
>> Note that slirp_lookup() converts VLAN to an int (without error
>> checking, i.e. non-integers are silently interpreted as zero, integers
>> out of bounds are silently truncated).
>>
>> Anyway, I can't make it work with -netdev:
>>
>> (qemu) hostfwd_add -1 net.0 tcp::12345-:22
>> unknown VLAN -1
>>
>> * Just for giggles: what if I give two arguments?
>>
>> $ qemu-system-x86_64--nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -net user -device virtio-net-pci,id=nic.0,vlan=0QEMU 0.15.92 monitor - type 'help' for more information
>> (qemu) hostfwd_add 0 user.0
>> invalid host forwarding rule '(null)'
>>
>> The (null) is from a printf-like function printing a null pointer with
>> conversion %s. Crash bug on some systems.
>>
>
> Agreed, there is some need for improvements. I would not mind changing
> the command syntax, in fact, if it helps and no major user is already
> relying on it (can't imagine, specifically as netdev support is broken).
> What about
>
> hostfwd_add rule [name [vlan]]
>
> ?
Should work quite nicely, as long as the targeted SlirpState has a name.
Always the case, isn't it?
Libvirt doesn't depend on hostfwd_add, as far as I can tell.
hostwd_remove should be changed to match.
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [Qemu-devel] hostfwd_add and -netdev user
2011-11-18 13:07 ` Markus Armbruster
@ 2011-11-18 13:13 ` Jan Kiszka
0 siblings, 0 replies; 4+ messages in thread
From: Jan Kiszka @ 2011-11-18 13:13 UTC (permalink / raw)
To: Markus Armbruster; +Cc: Mark McLoughlin, qemu-devel
[-- Attachment #1: Type: text/plain, Size: 3895 bytes --]
On 2011-11-18 11:07, Markus Armbruster wrote:
> Jan Kiszka <jan.kiszka@web.de> writes:
>
>> On 2011-11-16 08:07, Markus Armbruster wrote:
>>> Maybe I'm missing something, but it looks like hostfwd_add doesn't fully
>>> work with -netdev.
>>>
>>> Command definition from hmp-commands.hx:
>>>
>>> {
>>> .name = "hostfwd_add",
>>> .args_type = "arg1:s,arg2:s?,arg3:s?",
>>> .params = "[vlan_id name] [tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport",
>>> .help = "redirect TCP or UDP connections from host to guest (requires -net user)",
>>> .mhandler.cmd = net_slirp_hostfwd_add,
>>> },
>>>
>>> Note: params says command takes either 1 or 3 arguments. args_type
>>> can't express that, so it asks for 1-3.
>>>
>>> Command handler:
>>>
>>> void net_slirp_hostfwd_add(Monitor *mon, const QDict *qdict)
>>> {
>>> const char *redir_str;
>>> SlirpState *s;
>>> const char *arg1 = qdict_get_str(qdict, "arg1");
>>> const char *arg2 = qdict_get_try_str(qdict, "arg2");
>>> const char *arg3 = qdict_get_try_str(qdict, "arg3");
>>>
>>> if (arg2) {
>>> s = slirp_lookup(mon, arg1, arg2);
>>> redir_str = arg3;
>>> } else {
>>> s = slirp_lookup(mon, NULL, NULL);
>>> redir_str = arg1;
>>> }
>>> if (s) {
>>> slirp_hostfwd(s, redir_str, 0);
>>> }
>>>
>>> }
>>>
>>> If I understand the code correctly, the command has the following forms:
>>>
>>> * hostfwd_add REDIR
>>>
>>> This form uses the default stack (first member of slirp_stacks).
>>> Seems to work okay with -netdev:
>>>
>>> $ qemu-system-x86_64 --nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -netdev user,id=net.0 -device virtio-net-pci,netdev=net.0,id=nic.0
>>> QEMU 0.15.92 monitor - type 'help' for more information
>>> (qemu) hostfwd_add tcp::12345-:22
>>> (qemu) info usernet
>>> VLAN -1 (net.0):
>>> Protocol[State] FD Source Address Port Dest. Address Port RecvQ SendQ
>>> TCP[HOST_FORWARD] 12 * 12345 10.0.2.15 22 0 0
>>>
>>> * hostfwd_add VLAN STACK REDIR
>>>
>>> This form was added in commit f13b572c "slirp: Make hostfwd_add/remove
>>> multi-instance-aware".
>>>
>>> I don't like how it overloads the command; adding the optional
>>> arguments at the end would have been cleaner. Water under the bridge.
>>>
>>> Note that slirp_lookup() converts VLAN to an int (without error
>>> checking, i.e. non-integers are silently interpreted as zero, integers
>>> out of bounds are silently truncated).
>>>
>>> Anyway, I can't make it work with -netdev:
>>>
>>> (qemu) hostfwd_add -1 net.0 tcp::12345-:22
>>> unknown VLAN -1
>>>
>>> * Just for giggles: what if I give two arguments?
>>>
>>> $ qemu-system-x86_64--nodefaults --enable-kvm -vnc :0 -S -m 384 -monitor stdio -net user -device virtio-net-pci,id=nic.0,vlan=0QEMU 0.15.92 monitor - type 'help' for more information
>>> (qemu) hostfwd_add 0 user.0
>>> invalid host forwarding rule '(null)'
>>>
>>> The (null) is from a printf-like function printing a null pointer with
>>> conversion %s. Crash bug on some systems.
>>>
>>
>> Agreed, there is some need for improvements. I would not mind changing
>> the command syntax, in fact, if it helps and no major user is already
>> relying on it (can't imagine, specifically as netdev support is broken).
>> What about
>>
>> hostfwd_add rule [name [vlan]]
>>
>> ?
>
> Should work quite nicely, as long as the targeted SlirpState has a name.
> Always the case, isn't it?
IIRC, all network clients get model.instance as default name.
>
> Libvirt doesn't depend on hostfwd_add, as far as I can tell.
>
> hostwd_remove should be changed to match.
Yep.
Jan
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2011-11-18 13:14 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-11-16 10:07 [Qemu-devel] hostfwd_add and -netdev user Markus Armbruster
2011-11-18 12:37 ` Jan Kiszka
2011-11-18 13:07 ` Markus Armbruster
2011-11-18 13:13 ` Jan Kiszka
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).