Re: at smsc bug

ok it was in fact a bug. If no modem definition was used it was accessing memory which was never allocated.
SVN now has a fix for it.


On 06.08.2012, at 18:49, "Rene Kluwen" <[email protected]> wrote:

> I could “fix” it by changing line 252 of smsc_at.c to:
>  
>     if(privdata->modem && privdata->modem->hardware_flow_control) {
>  
> But… This assumes that if there’s still no modem detected at that time, hardware flow control is disabled.
> If that is not a correct assumption, the expression needs to be inverted.
>  
> == Rene
>  
> From: Rene Kluwen [mailto:[email protected]] 
> Sent: Monday, 06 August, 2012 18:46
> To: 'Rene Kluwen'; 'Andreas Fink'; [email protected]
> Subject: RE: at smsc bug
>  
> Output from gdb:
>  
> at2_open_device (privdata=0x9d712e0) at gw/smsc/smsc_at.c:252
> 252         if(privdata->modem->hardware_flow_control) {
> (gdb) backtrace
> #0  at2_open_device (privdata=0x9d712e0) at gw/smsc/smsc_at.c:252
> #1  0x0808764d in at2_test_speed (privdata=0x0, speed=5) at gw/smsc/smsc_at.c:2595
> #2  0x080877ca in at2_device_thread (arg=0x9d75240) at gw/smsc/smsc_at.c:2577
> #3  0x080d3541 in new_thread (arg=0x9d71420) at gwlib/gwthread-pthread.c:362
> #4  0x009bc45b in start_thread () from /lib/libpthread.so.0
> #5  0x0091423e in clone () from /lib/libc.so.6
>  
> From: Rene Kluwen [mailto:[email protected]] 
> Sent: Monday, 06 August, 2012 18:38
> To: 'Rene Kluwen'; 'Andreas Fink'; [email protected]
> Subject: RE: at smsc bug
>  
> Could it be that privdata->modem still has a NULL value at that time?
>  
> =+= Rene
>  
>  
> From: Rene Kluwen [mailto:[email protected]] 
> Sent: Monday, 06 August, 2012 18:32
> To: 'Rene Kluwen'; 'Andreas Fink'; [email protected]
> Subject: RE: at smsc bug
>  
> If I comment out the lines between:
>  
> #if defined(CRTSCTS) and #endif, then it works.
>  
> Weird…
>  
> == Rene
>  
> From: [email protected] [mailto:[email protected]] On Behalf Of Rene Kluwen
> Sent: Monday, 06 August, 2012 18:21
> To: 'Andreas Fink'; [email protected]
> Subject: RE: at smsc bug
>  
> These are the changes that were made since the version that worked (see attached document).
> I tried both hardware-flow-control = true and false. Both gave me the same error.
>  
> == Rene
>  
>  
> From: Andreas Fink [mailto:[email protected]] 
> Sent: Monday, 06 August, 2012 18:07
> To: Rene Kluwen
> Subject: Re: at smsc bug
>  
> please run it in with log-level=0 and send the output of the logfile so we can see the backtrace of the issue.
>  
> On 06.08.2012, at 18:01, Rene Kluwen <[email protected]> wrote:
>  
> 
> When starting current head revision in combination with a huwaei usb modem, I get the following.
> An older version of bearerbox has been working perfectly for some months already. This happened when I tried to upgrade.
> This is on a production machine, so it is a little bit hard to debug. The only thing that left me was downgrading again.
>  
> Logs:
>  
> 2012-08-06 17:55:07 [6608] [0] INFO: ----------------------------------------
> 2012-08-06 17:55:07 [6608] [0] INFO: Kannel bearerbox II version svn-r4988 starting
> 2012-08-06 17:55:07 [6608] [0] INFO: MAIN: Start-up done, entering mainloop
> 2012-08-06 17:55:07 [6608] [0] DEBUG: AT2[hvgsmsc]: start called
> 2012-08-06 17:55:07 [6608] [6] DEBUG: Thread 6 (gw/smsc/smsc_at.c:at2_device_thread) maps to pid 6608.
> 2012-08-06 17:55:07 [6608] [6] DEBUG: AT2[hvgsmsc]: detecting modem speed.
> 2012-08-06 17:55:07 [6608] [6] INFO: AT2[hvgsmsc]: opening device
> 2012-08-06 17:55:07 [6608] [6] DEBUG: AT2[hvgsmsc]: device opened. Telnet mode = 0
> 2012-08-06 17:55:07 [6608] [6] PANIC: ./bearerbox [0x80e8ebd]
> 2012-08-06 17:55:07 [6608] [6] PANIC: [0x160440]
> 2012-08-06 17:55:07 [6608] [6] PANIC: ./bearerbox [0x808764d]
> 2012-08-06 17:55:07 [6608] [6] PANIC: ./bearerbox [0x80877ca]
> 2012-08-06 17:55:07 [6608] [6] PANIC: ./bearerbox [0x80d3541]
> 2012-08-06 17:55:07 [6608] [6] PANIC: /lib/libpthread.so.0 [0x9bc45b]
> 2012-08-06 17:55:07 [6608] [6] PANIC: /lib/libc.so.6(clone+0x5e) [0x91423e]
> Segmentation fault
>  
> Config:
>  
> group = smsc
> smsc-id = hvgsmsc
> smsc = at
> modemtype=auto
> device=/dev/ttyUSB0
> my-number=+316xxxxxxx
> pin = 0000
> connect-allow-ip = 127.0.0.1
> sms-center = "+316540881000"
> log-level = 0
> sim-buffering = true
>  
> group = modems
> id = huawei
> name = "Huawei"
> detect-string = "huawei"
> detect-string2 = "K3765"
> init-string = "ATQ0 V1 E1 S0=0 &C1 &D2 +FCLASS=0; +CNMI=1,1,0,2,0; +CMEE=1"
> message-storage = "SM"
> message-start = 0
>  
>