8:41 p.m.
From: Martin Xu <martin.xu(a)intel.com>
Tramsmit ACCM and receive ACCM is mixed up.
According to RFC1662 Section 7.1, ACCM Configuration Option is
used to inform the peer which control characters MUST remain
mapped when the peer sends them.
---
gatchat/ppp_lcp.c | 16 ++++++++++++++--
1 files changed, 14 insertions(+), 2 deletions(-)
diff --git a/gatchat/ppp_lcp.c b/gatchat/ppp_lcp.c
index 3a80a62..cc3e231 100644
--- a/gatchat/ppp_lcp.c
+++ b/gatchat/ppp_lcp.c
@@ -149,7 +149,13 @@ static void lcp_rca(struct pppcp_data *pppcp, const struct
pppcp_packet *packet)
while (ppp_option_iter_next(&iter) == TRUE) {
switch (ppp_option_iter_get_type(&iter)) {
case ACCM:
- ppp_set_xmit_accm(pppcp_get_ppp(pppcp), 0);
+ /*
+ * RFC1662 Section 7.1
+ * The Configuration Option is used to inform the peer
+ * which control characters MUST remain mapped when
+ * the peer sends them.
+ */
+ ppp_set_recv_accm(pppcp_get_ppp(pppcp), 0);
break;
default:
break;
@@ -263,7 +269,13 @@ static enum rcr_result lcp_rcr(struct pppcp_data *pppcp,
while (ppp_option_iter_next(&iter) == TRUE) {
switch (ppp_option_iter_get_type(&iter)) {
case ACCM:
- ppp_set_recv_accm(ppp,
+ /*
+ * RFC1662 Section 7.1
+ * The Configuration Option is used to inform the peer
+ * which control characters MUST remain mapped when
+ * the peer sends them.
+ */
+ ppp_set_xmit_accm(ppp,
get_host_long(ppp_option_iter_get_data(&iter)));
break;
case AUTH_PROTO:
--
1.6.1.3
8:41 p.m.
New subject: [V3 PATCH 2/3] PPP: Use default ACCM (0xffffffff) to trasmit package
From: Martin Xu <martin.xu(a)intel.com>
Using my Huawei EM770W modem, if set ACCM as 0x00000000, RXJ-
event breaks PPP link, after IP package transmit for a while.
Using default ACCM, the issue can be fixed.
I tested it at China Unicom networks.
---
gatchat/ppp_lcp.c | 8 +++++---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/gatchat/ppp_lcp.c b/gatchat/ppp_lcp.c
index cc3e231..91a8b67 100644
--- a/gatchat/ppp_lcp.c
+++ b/gatchat/ppp_lcp.c
@@ -105,8 +105,7 @@ static void lcp_generate_config_options(struct lcp_data *lcp)
static void lcp_reset_config_options(struct lcp_data *lcp)
{
- lcp->req_options = REQ_OPTION_ACCM;
- lcp->accm = 0;
+ /* Using the default ACCM */
lcp_generate_config_options(lcp);
}
@@ -147,6 +146,7 @@ static void lcp_rca(struct pppcp_data *pppcp, const struct
pppcp_packet *packet)
ppp_option_iter_init(&iter, packet);
while (ppp_option_iter_next(&iter) == TRUE) {
+ const guint8 *data = ppp_option_iter_get_data(&iter);
switch (ppp_option_iter_get_type(&iter)) {
case ACCM:
/*
@@ -155,7 +155,9 @@ static void lcp_rca(struct pppcp_data *pppcp, const struct
pppcp_packet *packet)
* which control characters MUST remain mapped when
* the peer sends them.
*/
- ppp_set_recv_accm(pppcp_get_ppp(pppcp), 0);
+
+ ppp_set_recv_accm(pppcp_get_ppp(pppcp),
+ pppcpget_host_long(data));
break;
default:
break;
--
1.6.1.3
8:41 p.m.
New subject: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
From: Martin Xu <martin.xu(a)intel.com>
Device 00 is ppp port. Setting it as 01, my Huawei EM770W modem PPP
connection can't work
---
plugins/ofono.rules | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/plugins/ofono.rules b/plugins/ofono.rules
index 111f071..7dec243 100644
--- a/plugins/ofono.rules
+++ b/plugins/ofono.rules
@@ -37,7 +37,7 @@ ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1401",
ENV{OFONO_IFACE_NUM}=="02", E
ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1402",
ENV{OFONO_IFACE_NUM}=="00", ENV{OFONO_HUAWEI_TYPE}="Modem"
ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1402",
ENV{OFONO_IFACE_NUM}=="02", ENV{OFONO_HUAWEI_TYPE}="Pcui"
-ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1404",
ENV{OFONO_IFACE_NUM}=="01", ENV{OFONO_HUAWEI_TYPE}="Modem"
+ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1404",
ENV{OFONO_IFACE_NUM}=="00", ENV{OFONO_HUAWEI_TYPE}="Modem"
ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1404",
ENV{OFONO_IFACE_NUM}=="02", ENV{OFONO_HUAWEI_TYPE}="Pcui"
ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1404",
ENV{OFONO_HUAWEI_VOICE}="1"
--
1.6.1.3
7:28 a.m.
New subject: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
Hi Martin,
On 02/12/2011 03:41 AM, martin.xu(a)intel.com wrote:
From: Martin Xu <martin.xu(a)intel.com>
Device 00 is ppp port. Setting it as 01, my Huawei EM770W modem PPP
connection can't work
---
plugins/ofono.rules | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
This contradicts commit bb58a729073abcb6c5b0b2fb8b120973dfa42a07.
We really need a proper fix for this.
Regards,
-Denis
6:03 p.m.
New subject: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
Denis/Marcel:
-----Original Message-----
From: Denis Kenzior [mailto:denkenz@gmail.com]
Sent: Tuesday, February 15, 2011 4:28 AM
To: ofono(a)ofono.org
Cc: Xu, Martin
Subject: Re: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
Hi Martin,
On 02/12/2011 03:41 AM, martin.xu(a)intel.com wrote:
> From: Martin Xu <martin.xu(a)intel.com>
>
> Device 00 is ppp port. Setting it as 01, my Huawei EM770W modem PPP
> connection can't work
> ---
> plugins/ofono.rules | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
This contradicts commit bb58a729073abcb6c5b0b2fb8b120973dfa42a07.
I can't quite
understand why Dietrich needs to this commit. We formal configure can just make EM770W
works.
Could you help on this?
5:04 a.m.
New subject: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
Hi Martin,
On 02/15/2011 01:03 AM, Xu, Martin wrote:
Denis/Marcel:
> -----Original Message-----
> From: Denis Kenzior [mailto:denkenz@gmail.com]
> Sent: Tuesday, February 15, 2011 4:28 AM
> To: ofono(a)ofono.org
> Cc: Xu, Martin
> Subject: Re: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
>
> Hi Martin,
>
> On 02/12/2011 03:41 AM, martin.xu(a)intel.com wrote:
>> From: Martin Xu <martin.xu(a)intel.com>
>>
>> Device 00 is ppp port. Setting it as 01, my Huawei EM770W modem PPP
>> connection can't work
>> ---
>> plugins/ofono.rules | 2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>
>
> This contradicts commit bb58a729073abcb6c5b0b2fb8b120973dfa42a07.
I can't quite understand why Dietrich needs to this commit. We formal configure can
just make EM770W works.
Could you help on this?
The Huawei firmware is funny, it can actually reconfigure the modem and
pcui ports (e.g. swap them). So we might have to be a bit smarter and
figure out which ports are actually being used for what inside the modem
driver.
Talk to Marcel, I believe he has played with the Huawei hardware the
most and might have ideas on how to fix this.
Regards,
-Denis
1:19 p.m.
New subject: [V3 PATCH 3/3] Huawei: set Huawei EM770W modem device to 00
Hi Denis,
>> This contradicts commit
bb58a729073abcb6c5b0b2fb8b120973dfa42a07.
> I can't quite understand why Dietrich needs to this commit. We formal configure
can just make EM770W works.
> Could you help on this?
The Huawei firmware is funny, it can actually reconfigure the modem and
pcui ports (e.g. swap them). So we might have to be a bit smarter and
figure out which ports are actually being used for what inside the modem
driver.
so I did apply this patch now since in the default settings the MDM port
should be on interface 0. All other ones are tempered with.
Talk to Marcel, I believe he has played with the Huawei hardware the
most and might have ideas on how to fix this.
And yes, I will add auto-detection for this now. It should be be
possible to figure this out.
ofonod[908]: PCUI: > AT^GETPORTMODE\r
ofonod[908]: PCUI: <
\r\n^getportmode:type:WCDMA:Qualcomm,MDM:0,NDIS:1,DIAG:2,PCUI:3\r\n\r\nOK\r\n
All latest modules support ^GETPORTMODE command and the ones that don't
have only two ports anyway.
Regards
Marcel
7:26 a.m.
New subject: [V3 PATCH 2/3] PPP: Use default ACCM (0xffffffff) to trasmit package
Hi Martin,
On 02/12/2011 03:41 AM, martin.xu(a)intel.com wrote:
From: Martin Xu <martin.xu(a)intel.com>
Using my Huawei EM770W modem, if set ACCM as 0x00000000, RXJ-
event breaks PPP link, after IP package transmit for a while.
Using default ACCM, the issue can be fixed.
I tested it at China Unicom networks.
---
gatchat/ppp_lcp.c | 8 +++++---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/gatchat/ppp_lcp.c b/gatchat/ppp_lcp.c
index cc3e231..91a8b67 100644
--- a/gatchat/ppp_lcp.c
+++ b/gatchat/ppp_lcp.c
@@ -105,8 +105,7 @@ static void lcp_generate_config_options(struct lcp_data *lcp)
static void lcp_reset_config_options(struct lcp_data *lcp)
{
- lcp->req_options = REQ_OPTION_ACCM;
- lcp->accm = 0;
+ /* Using the default ACCM */
lcp_generate_config_options(lcp);
}
@@ -147,6 +146,7 @@ static void lcp_rca(struct pppcp_data *pppcp, const struct
pppcp_packet *packet)
ppp_option_iter_init(&iter, packet);
while (ppp_option_iter_next(&iter) == TRUE) {
+ const guint8 *data = ppp_option_iter_get_data(&iter);
switch (ppp_option_iter_get_type(&iter)) {
case ACCM:
/*
@@ -155,7 +155,9 @@ static void lcp_rca(struct pppcp_data *pppcp, const struct
pppcp_packet *packet)
* which control characters MUST remain mapped when
* the peer sends them.
*/
- ppp_set_recv_accm(pppcp_get_ppp(pppcp), 0);
+
+ ppp_set_recv_accm(pppcp_get_ppp(pppcp),
+ pppcpget_host_long(data));
make --no-print-directory all-am
CC gatchat/ppp_lcp.o
cc1: warnings being treated as errors
gatchat/ppp_lcp.c: In function ‘lcp_rca’:
gatchat/ppp_lcp.c:160: error: implicit declaration of function
‘pppcpget_host_long’
make[1]: *** [gatchat/ppp_lcp.o] Error 1
make: *** [all] Error 2
break;
default:
break;
Regards,
-Denis
7:25 a.m.
Hi Martin,
On 02/12/2011 03:41 AM, martin.xu(a)intel.com wrote:
From: Martin Xu <martin.xu(a)intel.com>
Tramsmit ACCM and receive ACCM is mixed up.
According to RFC1662 Section 7.1, ACCM Configuration Option is
used to inform the peer which control characters MUST remain
mapped when the peer sends them.
---
gatchat/ppp_lcp.c | 16 ++++++++++++++--
1 files changed, 14 insertions(+), 2 deletions(-)
Patch has been applied, thanks.
Regards,
-Denis
4231
days inactive
4236
days old
8 comments
4 participants
participants (4)
-
Denis Kenzior -
Marcel Holtmann -
martin.xu＠intel.com -
Xu, Martin