Optware cups segfaults on TS-119

[Kay Hannay-2]

Hi,

I have installed the cupsd package from optware on my Qnap TS-119.
When I try to start cupsd, I just get a segmentation fault. The same
for other cups related binaries like accept, cupsctl, lpadmin etc.
I have the new Firmware version 3.0.0 Build 0430T on my TS-119.

Here are the last lines of strace after starting cupsd:

open("/opt/lib/libsasl2.so.2", O_RDONLY) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0\3500\0\0004\0\0\0"..., 512) = 512
fstat64(3, {st_mode=S_IFREG|0755, st_size=86212, ...}) = 0
mmap2(NULL, 117872, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x4049f000
mprotect(0x404b4000, 28672, PROT_NONE)  = 0
mmap2(0x404bb000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x14) = 0x404bb000
close(3)                                = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x404bc000
syscall_983045(0x404bc7f0, 0x400256ec, 0x404bcec8, 0x40025050, 0x40, 0x40025000, 0, 0xf0005, \
    0x40025730, 0x1, 0, 0xbefdc55c, 0, 0xbefdc540, 0x88, 0x4000221c, 0x20000010, 0x404bc7f0, \
    0, 0, 0xdb1c, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) = 0
mprotect(0x4049b000, 4096, PROT_READ)   = 0
mprotect(0x40470000, 4096, PROT_READ)   = 0
mprotect(0x40460000, 4096, PROT_READ)   = 0
mprotect(0x40309000, 4096, PROT_READ)   = 0
mprotect(0x402fb000, 4096, PROT_READ)   = 0
mprotect(0x40249000, 4096, PROT_READ)   = 0
munmap(0x4001f000, 17045)               = 0
set_tid_address(0x404bc398)             = 12577
SYS_338(0x404bc3a0, 0xc, 0x404bc7f0, 0xffffffec, 0x4001e400) = 0
rt_sigaction(SIGRTMIN, {0x40230f74, [], SA_SIGINFO|0x4000000}, NULL, 8) = 0
rt_sigaction(SIGRT_1, {0x40230e4c, [], SA_RESTART|SA_SIGINFO|0x4000000}, NULL, 8) = 0
rt_sigprocmask(SIG_UNBLOCK, [RTMIN RT_1], NULL, 8) = 0
getrlimit(RLIMIT_STACK, {rlim_cur=8192*1024, rlim_max=RLIM_INFINITY}) = 0
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++

Does anybody have the same behaviour with cups?

Thanks!
 
 
Best regards
Kay

[brian_zhou]

Please re-test with 1.3.10-2.

I made the change to turn off stack-protector in gcc compilation option.

Regards,

-Brian

--- In nslu2-linux@..., "Kay Hannay" <mskbike@...> wrote:

>
> Hi,
>
> I have installed the cupsd package from optware on my Qnap TS-119.
> When I try to start cupsd, I just get a segmentation fault. The same
> for other cups related binaries like accept, cupsctl, lpadmin etc.
> I have the new Firmware version 3.0.0 Build 0430T on my TS-119.
>
> Here are the last lines of strace after starting cupsd:
>
> open("/opt/lib/libsasl2.so.2", O_RDONLY) = 3
> read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0\3500\0\0004\0\0\0"..., 512) = 512
> fstat64(3, {st_mode=S_IFREG|0755, st_size=86212, ...}) = 0
> mmap2(NULL, 117872, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x4049f000
> mprotect(0x404b4000, 28672, PROT_NONE)  = 0
> mmap2(0x404bb000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x14) = 0x404bb000
> close(3)                                = 0
> mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x404bc000
> syscall_983045(0x404bc7f0, 0x400256ec, 0x404bcec8, 0x40025050, 0x40, 0x40025000, 0, 0xf0005, \
>     0x40025730, 0x1, 0, 0xbefdc55c, 0, 0xbefdc540, 0x88, 0x4000221c, 0x20000010, 0x404bc7f0, \
>     0, 0, 0xdb1c, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) = 0
> mprotect(0x4049b000, 4096, PROT_READ)   = 0
> mprotect(0x40470000, 4096, PROT_READ)   = 0
> mprotect(0x40460000, 4096, PROT_READ)   = 0
> mprotect(0x40309000, 4096, PROT_READ)   = 0
> mprotect(0x402fb000, 4096, PROT_READ)   = 0
> mprotect(0x40249000, 4096, PROT_READ)   = 0
> munmap(0x4001f000, 17045)               = 0
> set_tid_address(0x404bc398)             = 12577
> SYS_338(0x404bc3a0, 0xc, 0x404bc7f0, 0xffffffec, 0x4001e400) = 0
> rt_sigaction(SIGRTMIN, {0x40230f74, [], SA_SIGINFO|0x4000000}, NULL, 8) = 0
> rt_sigaction(SIGRT_1, {0x40230e4c, [], SA_RESTART|SA_SIGINFO|0x4000000}, NULL, 8) = 0
> rt_sigprocmask(SIG_UNBLOCK, [RTMIN RT_1], NULL, 8) = 0
> getrlimit(RLIMIT_STACK, {rlim_cur=8192*1024, rlim_max=RLIM_INFINITY}) = 0
> --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> +++ killed by SIGSEGV +++
>
> Does anybody have the same behaviour with cups?
>
> Thanks!
>  
>  
> Best regards
> Kay
>

[Kay Hannay-2]

Hi Brian,

I have updated the cups package and it seems to work now. I have to test later with my printer at home. I tried the cups-pdf printer, which does not work. But this seems to be related to the cups-pdf package itself.

I can connect to the web interface etc without problems.

Thank you!!

Best regards
Kay

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:

>
> Please re-test with 1.3.10-2.
>
> I made the change to turn off stack-protector in gcc compilation option.
>
> Regards,
>
> -Brian
>
> --- In nslu2-linux@..., "Kay Hannay" <mskbike@> wrote:
> >
> > Hi,
> >
> > I have installed the cupsd package from optware on my Qnap TS-119.
> > When I try to start cupsd, I just get a segmentation fault. The same
> > for other cups related binaries like accept, cupsctl, lpadmin etc.
> > I have the new Firmware version 3.0.0 Build 0430T on my TS-119.
> >
> > Here are the last lines of strace after starting cupsd:
> >
> > open("/opt/lib/libsasl2.so.2", O_RDONLY) = 3
> > read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0\3500\0\0004\0\0\0"..., 512) = 512
> > fstat64(3, {st_mode=S_IFREG|0755, st_size=86212, ...}) = 0
> > mmap2(NULL, 117872, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x4049f000
> > mprotect(0x404b4000, 28672, PROT_NONE)  = 0
> > mmap2(0x404bb000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x14) = 0x404bb000
> > close(3)                                = 0
> > mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x404bc000
> > syscall_983045(0x404bc7f0, 0x400256ec, 0x404bcec8, 0x40025050, 0x40, 0x40025000, 0, 0xf0005, \
> >     0x40025730, 0x1, 0, 0xbefdc55c, 0, 0xbefdc540, 0x88, 0x4000221c, 0x20000010, 0x404bc7f0, \
> >     0, 0, 0xdb1c, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) = 0
> > mprotect(0x4049b000, 4096, PROT_READ)   = 0
> > mprotect(0x40470000, 4096, PROT_READ)   = 0
> > mprotect(0x40460000, 4096, PROT_READ)   = 0
> > mprotect(0x40309000, 4096, PROT_READ)   = 0
> > mprotect(0x402fb000, 4096, PROT_READ)   = 0
> > mprotect(0x40249000, 4096, PROT_READ)   = 0
> > munmap(0x4001f000, 17045)               = 0
> > set_tid_address(0x404bc398)             = 12577
> > SYS_338(0x404bc3a0, 0xc, 0x404bc7f0, 0xffffffec, 0x4001e400) = 0
> > rt_sigaction(SIGRTMIN, {0x40230f74, [], SA_SIGINFO|0x4000000}, NULL, 8) = 0
> > rt_sigaction(SIGRT_1, {0x40230e4c, [], SA_RESTART|SA_SIGINFO|0x4000000}, NULL, 8) = 0
> > rt_sigprocmask(SIG_UNBLOCK, [RTMIN RT_1], NULL, 8) = 0
> > getrlimit(RLIMIT_STACK, {rlim_cur=8192*1024, rlim_max=RLIM_INFINITY}) = 0
> > --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> > +++ killed by SIGSEGV +++
> >
> > Does anybody have the same behaviour with cups?
> >
> > Thanks!
> >  
> >  
> > Best regards
> > Kay
> >
>

[Kay Hannay-2]

Hello again,

cups-pdf is also working now. There were a few things I had to do:

- copy cups-pdf.conf from /opt/etc/ to /opt/etc/cups/
- create the temp directory for gs (/opt/var/tmp) or set the variable GSTmp to a different directory in cups-pdf.conf
- set "Grp administrators" in cups-pdf.conf
- set the GhostScript binary to "GhostScript /opt/bin/gs" in cups-pdf.conf

Now I get proper PDF documents.


Regards
Kay

--- In nslu2-linux@..., "Kay Hannay" <mskbike@...> wrote:

>
> Hi Brian,
>
> I have updated the cups package and it seems to work now. I have to test later with my printer at home. I tried the cups-pdf printer, which does not work. But this seems to be related to the cups-pdf package itself.
>
> I can connect to the web interface etc without problems.
>
> Thank you!!
>
> Best regards
> Kay
>
> --- In nslu2-linux@..., "Brian Zhou" <b88zhou@> wrote:
> >
> > Please re-test with 1.3.10-2.
> >
> > I made the change to turn off stack-protector in gcc compilation option.
> >
> > Regards,
> >
> > -Brian
> >
> > --- In nslu2-linux@..., "Kay Hannay" <mskbike@> wrote:
> > >
> > > Hi,
> > >
> > > I have installed the cupsd package from optware on my Qnap TS-119.
> > > When I try to start cupsd, I just get a segmentation fault. The same
> > > for other cups related binaries like accept, cupsctl, lpadmin etc.
> > > I have the new Firmware version 3.0.0 Build 0430T on my TS-119.
> > >
> > > Here are the last lines of strace after starting cupsd:
> > >
> > > open("/opt/lib/libsasl2.so.2", O_RDONLY) = 3
> > > read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0(\0\1\0\0\0\3500\0\0004\0\0\0"..., 512) = 512
> > > fstat64(3, {st_mode=S_IFREG|0755, st_size=86212, ...}) = 0
> > > mmap2(NULL, 117872, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x4049f000
> > > mprotect(0x404b4000, 28672, PROT_NONE)  = 0
> > > mmap2(0x404bb000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x14) = 0x404bb000
> > > close(3)                                = 0
> > > mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x404bc000
> > > syscall_983045(0x404bc7f0, 0x400256ec, 0x404bcec8, 0x40025050, 0x40, 0x40025000, 0, 0xf0005, \
> > >     0x40025730, 0x1, 0, 0xbefdc55c, 0, 0xbefdc540, 0x88, 0x4000221c, 0x20000010, 0x404bc7f0, \
> > >     0, 0, 0xdb1c, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) = 0
> > > mprotect(0x4049b000, 4096, PROT_READ)   = 0
> > > mprotect(0x40470000, 4096, PROT_READ)   = 0
> > > mprotect(0x40460000, 4096, PROT_READ)   = 0
> > > mprotect(0x40309000, 4096, PROT_READ)   = 0
> > > mprotect(0x402fb000, 4096, PROT_READ)   = 0
> > > mprotect(0x40249000, 4096, PROT_READ)   = 0
> > > munmap(0x4001f000, 17045)               = 0
> > > set_tid_address(0x404bc398)             = 12577
> > > SYS_338(0x404bc3a0, 0xc, 0x404bc7f0, 0xffffffec, 0x4001e400) = 0
> > > rt_sigaction(SIGRTMIN, {0x40230f74, [], SA_SIGINFO|0x4000000}, NULL, 8) = 0
> > > rt_sigaction(SIGRT_1, {0x40230e4c, [], SA_RESTART|SA_SIGINFO|0x4000000}, NULL, 8) = 0
> > > rt_sigprocmask(SIG_UNBLOCK, [RTMIN RT_1], NULL, 8) = 0
> > > getrlimit(RLIMIT_STACK, {rlim_cur=8192*1024, rlim_max=RLIM_INFINITY}) = 0
> > > --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> > > +++ killed by SIGSEGV +++
> > >
> > > Does anybody have the same behaviour with cups?
> > >
> > > Thanks!
> > >  
> > >  
> > > Best regards
> > > Kay
> > >
> >
>

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> Please re-test with 1.3.10-2.
>
> I made the change to turn off stack-protector in gcc compilation option.
>
> Regards,
>
> -Brian

Hi Brian,

I have successfully tested on TS-x19 (119 & 219 model) cups and all work fine, even if we have to modify some path in foomatic-rip to add the /opt/bin and /opt/lib/cups/filter ...

I now try to do the same things with a TS-109 (previous series that used the feed http://ipkg.nslu2-linux.org/feeds/optware/cs06q3armel/cross/unstable/ and not the feed used by TS-x19 :  http://ipkg.nslu2-linux.org/feeds/optware/cs08q1armel/cross/unstable/

Or we have always seg. fault, so my question is :
did you do the same job for the cs06q3armel feed ?, to understand if it's another probleme or just a compilation diff.

Thanks a lot, great job, you really help us .

Philippe.

[brian_zhou]

IIRC TS-x09 should use cs05q1armel feed (not cs05q3armel), there's no cups package in cs06q3armel feed. The cups compilation option change affect all platforms with gcc4. The segfault you're having maybe a different problem.

-Brian

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:

>
> --- In nslu2-linux@..., "Brian Zhou" <b88zhou@> wrote:
> >
> > Please re-test with 1.3.10-2.
> >
> > I made the change to turn off stack-protector in gcc compilation option.
> >
> > Regards,
> >
> > -Brian
>
> Hi Brian,
>
> I have successfully tested on TS-x19 (119 & 219 model) cups and all work fine, even if we have to modify some path in foomatic-rip to add the /opt/bin and /opt/lib/cups/filter ...
>
> I now try to do the same things with a TS-109 (previous series that used the feed http://ipkg.nslu2-linux.org/feeds/optware/cs06q3armel/cross/unstable/ and not the feed used by TS-x19 :  http://ipkg.nslu2-linux.org/feeds/optware/cs08q1armel/cross/unstable/
>
> Or we have always seg. fault, so my question is :
> did you do the same job for the cs06q3armel feed ?, to understand if it's another probleme or just a compilation diff.
>
> Thanks a lot, great job, you really help us .
>
> Philippe.
>

[brian_zhou]

On another note, please try document any configuration file changes you need to make. Ideally as a diff/patch to optware/make/cups.mk, but at least on a wiki page. This is so that it can be integrated in the next build.

Thanks,

-Brian

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:

>
> --- In nslu2-linux@..., "Brian Zhou" <b88zhou@> wrote:
> >
> > Please re-test with 1.3.10-2.
> >
> > I made the change to turn off stack-protector in gcc compilation option.
> >
> > Regards,
> >
> > -Brian
>
> Hi Brian,
>
> I have successfully tested on TS-x19 (119 & 219 model) cups and all work fine, even if we have to modify some path in foomatic-rip to add the /opt/bin and /opt/lib/cups/filter ...
>
> I now try to do the same things with a TS-109 (previous series that used the feed http://ipkg.nslu2-linux.org/feeds/optware/cs06q3armel/cross/unstable/ and not the feed used by TS-x19 :  http://ipkg.nslu2-linux.org/feeds/optware/cs08q1armel/cross/unstable/
>
> Or we have always seg. fault, so my question is :
> did you do the same job for the cs06q3armel feed ?, to understand if it's another probleme or just a compilation diff.
>
> Thanks a lot, great job, you really help us .
>
> Philippe.
>

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> IIRC TS-x09 should use cs05q1armel feed (not cs05q3armel), there's no cups package in cs06q3armel feed. The cups compilation option change affect all platforms with gcc4. The segfault you're having maybe a different problem.
>
> -Brian
>
Oups!!! sorry Brian I do an error ... a buggy cut & paste ...

so I think it's another bug ... I will try to understand ... for your information
we can't run web interface, at first access cupsd fall in seg. fault ...
but by hand, we are able to define printer and use it ... only no syntax fault are authorize ... but I arrive to print the cups test page at hand to a HP thru network ... so not so bad.

Philippe.

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> On another note, please try document any configuration file changes you need to make. Ideally as a diff/patch to optware/make/cups.mk, but at least on a wiki page. This is so that it can be integrated in the next build.
>
> Thanks,
>
> -Brian
>
Hi Brian,
Not sure that I am able to propose a modify cups.mk ... but at the moment I write a how-to based on my experience on TS-x19 and TS-x09 series with all the modification to do ... so I will resume in a mess ...all the points, I am sure for some of them ... but for some other I do some test and keep value, even if it's due to another things ...

But I will write the how-to, at first on qnap forum.

Thanks again.
Philippe.

[Kay Hannay-2]

Hallo,

--- In nslu2-linux@..., "Kay Hannay" <mskbike@...> wrote:
> I have updated the cups package and it seems to work now. I have to
> test later with my printer at home.

as promised, I tested cups on the TS-119 with a real USB printer (Canon i560).
I had to change the rights to the /dev/usb/lp* devices to 666 and to install
the gutenprint package, which was not installed automatically by ipkg.
After that the printer is recognized by cups and I can select it from the web
interface. Printing works fine.

So thank you very much, Brian!


Best regards
Kay

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> IIRC TS-x09 should use cs05q1armel feed (not cs05q3armel), there's no cups package in cs06q3armel feed. The cups compilation option change affect all platforms with gcc4. The segfault you're having maybe a different problem.
>
> -Brian
>
Hi Brian (with the good feed ....) I do a strace of cupsd on TS-109 if you have an idea for this seg. fault ....
I just put the strace after cupsd is waiting for connection, but I can provide a full one.
Please just have a look if you see something, but don't spend more time ...

Thanks Philippe.

epoll_wait(0x3, 0x607e8, 0x400, 0x3e8)  = 0
time(NULL)                              = 1246110900
time(NULL)                              = 1246110900
time(NULL)                              = 1246110900
time(NULL)                              = 1246110900
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 49) = 49
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 46) = 46
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 53) = 53
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 50) = 50
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 59) = 59
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 67) = 67
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 67) = 67
time(NULL)                              = 1246110900
write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 67) = 67
time(NULL)                              = 1246110900
epoll_wait(0x3, 0x607e8, 0x400, 0xffffffff) = 1
time(NULL)                              = 1246110922
accept(4, {sa_family=AF_INET, sin_port=htons(35423), sin_addr=inet_addr("127.0.0.1")}, [16]) = 11
time(NULL)                              = 1246110922
write(5, "D [27/Jun/2009:15:55:22 +0200] c"..., 79) = 79
getsockname(11, {sa_family=AF_INET, sin_port=htons(631), sin_addr=inet_addr("127.0.0.1")}, [16]) = 0
setsockopt(11, SOL_TCP, TCP_NODELAY, [1], 4) = 0
fcntl64(11, F_GETFD)                    = 0
fcntl64(11, F_SETFD, FD_CLOEXEC)        = 0
epoll_ctl(0x3, 0x1, 0xb, 0xbeffa008)    = 0
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
epoll_wait(0x3, 0x607e8, 0x400, 0x3e8)  = 2
recv(11, "G", 1, MSG_PEEK)              = 1
poll([{fd=11, events=POLLIN, revents=POLLIN}], 1, 10000) = 1
recv(11, "GET /images/top-left.gif HTTP/1."..., 2048, 0) = 431
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
write(5, "D [27/Jun/2009:15:55:22 +0200] c"..., 85) = 85
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
access("/opt/share/locale/fr/cups_fr.po", F_OK) = 0
open("/opt/share/locale/fr/cups_fr.po", O_RDONLY|O_LARGEFILE) = 12
fcntl64(12, F_GETFD)                    = 0
fcntl64(12, F_SETFD, FD_CLOEXEC)        = 0
read(12, "msgid \"\\t\\t(all)\\n\"\nmsgstr \"\\t\\t"..., 4096) = 4096
read(12, "      PASS    LanguageVersion\\n\""..., 4096) = 4096
read(12, "  ShortNickName required by PPD "..., 4096) = 4096
read(12, "r than 31 chars.\\n              "..., 4096) = 4096
read(12, "PSVersion\\n                REF: "..., 4096) = 4096
read(12, "**FAIL**  Unable to open PPD fil"..., 4096) = 4096
read(12, "\231option \302\253\302\240-o\302\240\302\273.\\n\"\nmsgid \"%s"..., 4096) = 4096
read(12, " ignored!\\n\"\nmsgstr \"%s: Attenti"..., 4096) = 4096
read(12, "mat \302\253 %s \302\273.\"\nmsgid \"Cover open"..., 4096) = 4096
read(12, "\"ERROR: l\342\200\231h\303\264te \303\240 distance n\342"..., 4096) = 4096
read(12, "R: Unknown format character \\\"%c"..., 4096) = 4096
read(12, " d\342\200\231impression...\\n\"\nmsgid \"INF"..., 4096) = 4096
read(12, " )\"\nmsgid \"Job #%d cannot be res"..., 4096) = 4096
read(12, " pointer\"\nmsgstr \"Pointeur de fi"..., 4096) = 4096
read(12, "\nmsgstr \"D\303\251finir les autorisati"..., 4096) = 4096
read(12, "ble to copy Windows 2000 printer"..., 4096) = 4096
read(12, "           [-r remove-class] [-v"..., 4096) = 4096
read(12, " [ options ] nom de fichier(s)\\n"..., 4096) = 4096
read(12, "a pas r\303\251pondu avec un octet d\342\200"..., 4096) = 4096
read(12, "\302\240\302\273.\"\nmsgid \"lpadmin: Class nam"..., 4096) = 4096
read(12, "to set the device URI:\\n        "..., 4096) = 4096
brk(0xc2000)                            = 0xc2000
read(12, "gid \"lppasswd: Unable to open pa"..., 4096) = 3186
read(12, "", 4096)                      = 0
close(12)                               = 0
time(NULL)                              = 1246110922
write(5, "D [27/Jun/2009:15:55:22 +0200] c"..., 80) = 80
stat64("/opt/share/doc/cups/fr/images/top-left.gif", 0xbefefac8) = -1 ENOENT (No such file or directory)
stat64("/opt/share/doc/cups/fr/images/top-left.gif", 0xbefefac8) = -1 ENOENT (No such file or directory)
stat64("/opt/share/doc/cups/images/top-left.gif", {st_mode=S_IFREG|0644, st_size=17592186044416, ...}) = 0
open("/opt/share/doc/cups/images/top-left.gif", O_RDONLY|O_LARGEFILE) = 12
fcntl64(12, F_GETFD)                    = 0
fcntl64(12, F_SETFD, FD_CLOEXEC)        = 0
read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
read(12, "", 4096)                      = 0
_llseek(12, 0, [0], SEEK_SET)           = 0
read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
read(12, "", 4096)                      = 0
_llseek(12, 0, [0], SEEK_SET)           = 0
read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
read(12, "", 4096)                      = 0
_llseek(12, 0, [0], SEEK_SET)           = 0
read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
read(12, "", 4096)                      = 0
close(12)                               = 0
open("/opt/share/doc/cups/images/top-left.gif", O_RDONLY|O_LARGEFILE) = 12
fcntl64(12, F_GETFD)                    = 0
fcntl64(12, F_SETFD, FD_CLOEXEC)        = 0
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
time(NULL)                              = 1246110922
send(11, "HTTP/1.1 200 OK\r\nDate: Sat, 27 J"..., 236, 0) = 236
epoll_ctl(0x3, 0x3, 0xb, 0xbefefb50)    = 0
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> On another note, please try document any configuration file changes you need to make. Ideally as a diff/patch to optware/make/cups.mk, but at least on a wiki page. This is so that it can be integrated in the next build.
>
> Thanks,
>
> -Brian
Hi brian ... a summary of change ...
I don't write about change need by "normal configuration step", like Listen or Allow directive, just some part able to be change in the script or in base config. files
in cups :
cupsd.conf .... add administrators to the list of group for lp
SystemGroup sys root administrators
so sys and root group keep valid but for TS-xxx administrators is the good one
in hplip (I suppose) ... the foomatic-rip perl script
so please add perl as a requirement for hplip.
foomatic-rip don't exist only foomatic-rip-hplip, please add a link to.
I don't know about hpijs ... because if foomatic-rip exist he fall during exec. (in fact it's gs, the same (gs) that work with hplip ?) ....
change in foomatic-rip, please modify the lines :
my $execpath = "NONE/bin:/usr/local/bin:/usr/bin:/bin:/opt/bin:/opt/sbin";.... add /opt/bin & /opt/sbin
my $cupsfilterpath = "/usr/local/lib/cups/filter:/usr/local/lib/cups/filter:/usr/local/libexec/cups/filter:/opt/lib/cups/filter:/usr/lib/cups/filter"; ...... add /opt/lib/cups/filter

for cups-pdf
put the cups-pdf.conf directly under /opt/etc/cups
change (if it's possible) or add the administrators group to lp group
#Grp lp
Grp administrators
and the path for gs to be coherent with other ipkg (ghostscript)
GhostScript /opt/bin/gs
/opt/var/tmp is mandatory but not created by ipkg install

I don't know if it's only for TS-x09 or if it's not a problem if we put them on other system ??? but TS-x09 needed sticky bit on cups-pdf backend to work.
[/opt/lib/cups/backend] # ll cups-pdf
-rwsr-sr-x 1 admin administ 22.4k Feb 14 17:27 cups-pdf*
I don't know if post-install is done during ipkg install
I have write a basic how-to (based on my own experience) for qnap TS-x09 & TS-x19 ... if you are interested to read ? :
http://forum.qnap.com/viewtopic.php?f=85&t=15686#p72101

Again thanks for the great job ...

Philippe.

[phild_mande]

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:
>
> I don't know if it's only for TS-x09 or if it's not a problem if we put them on other system ??? but TS-x09 needed sticky bit on cups-pdf backend to work.
> [/opt/lib/cups/backend] # ll cups-pdf
> -rwsr-sr-x 1 admin administ 22.4k Feb 14 17:27 cups-pdf*

Hi

Soory, I am tired ... it's the SUID bit NOT the sticky bit ... so sorry, even if I try always to do the file visibility .. so you can change by yourself.

Apologize.
Philippe.

[brian_zhou]

It is known from users of transmission, that a few platforms using cs05q1armel feed need to set EVENT_NOEPOLL with a non-empty value, to work around the (possibly kernel) problem.

But since cups does not use libevent, for this feed, probably better to disable epoll during configure with ac_cv_func_epoll_create=no (before line 189 of make/cups.mk)?  I don't have a device to test, someone should give it a try.

Good luck,

-Brian

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:

>
> --- In nslu2-linux@..., "Brian Zhou" <b88zhou@> wrote:
> >
> > IIRC TS-x09 should use cs05q1armel feed (not cs05q3armel), there's no cups package in cs06q3armel feed. The cups compilation option change affect all platforms with gcc4. The segfault you're having maybe a different problem.
> >
> > -Brian
> >
> Hi Brian (with the good feed ....) I do a strace of cupsd on TS-109 if you have an idea for this seg. fault ....
> I just put the strace after cupsd is waiting for connection, but I can provide a full one.
> Please just have a look if you see something, but don't spend more time ...
>
> Thanks Philippe.
>
> epoll_wait(0x3, 0x607e8, 0x400, 0x3e8)  = 0
> time(NULL)                              = 1246110900
> time(NULL)                              = 1246110900
> time(NULL)                              = 1246110900
> time(NULL)                              = 1246110900
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 49) = 49
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 46) = 46
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 53) = 53
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 50) = 50
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 59) = 59
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 67) = 67
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 67) = 67
> time(NULL)                              = 1246110900
> write(5, "D [27/Jun/2009:15:55:00 +0200] R"..., 67) = 67
> time(NULL)                              = 1246110900
> epoll_wait(0x3, 0x607e8, 0x400, 0xffffffff) = 1
> time(NULL)                              = 1246110922
> accept(4, {sa_family=AF_INET, sin_port=htons(35423), sin_addr=inet_addr("127.0.0.1")}, [16]) = 11
> time(NULL)                              = 1246110922
> write(5, "D [27/Jun/2009:15:55:22 +0200] c"..., 79) = 79
> getsockname(11, {sa_family=AF_INET, sin_port=htons(631), sin_addr=inet_addr("127.0.0.1")}, [16]) = 0
> setsockopt(11, SOL_TCP, TCP_NODELAY, [1], 4) = 0
> fcntl64(11, F_GETFD)                    = 0
> fcntl64(11, F_SETFD, FD_CLOEXEC)        = 0
> epoll_ctl(0x3, 0x1, 0xb, 0xbeffa008)    = 0
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> epoll_wait(0x3, 0x607e8, 0x400, 0x3e8)  = 2
> recv(11, "G", 1, MSG_PEEK)              = 1
> poll([{fd=11, events=POLLIN, revents=POLLIN}], 1, 10000) = 1
> recv(11, "GET /images/top-left.gif HTTP/1."..., 2048, 0) = 431
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> write(5, "D [27/Jun/2009:15:55:22 +0200] c"..., 85) = 85
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> access("/opt/share/locale/fr/cups_fr.po", F_OK) = 0
> open("/opt/share/locale/fr/cups_fr.po", O_RDONLY|O_LARGEFILE) = 12
> fcntl64(12, F_GETFD)                    = 0
> fcntl64(12, F_SETFD, FD_CLOEXEC)        = 0
> read(12, "msgid \"\\t\\t(all)\\n\"\nmsgstr \"\\t\\t"..., 4096) = 4096
> read(12, "      PASS    LanguageVersion\\n\""..., 4096) = 4096
> read(12, "  ShortNickName required by PPD "..., 4096) = 4096
> read(12, "r than 31 chars.\\n              "..., 4096) = 4096
> read(12, "PSVersion\\n                REF: "..., 4096) = 4096
> read(12, "**FAIL**  Unable to open PPD fil"..., 4096) = 4096
> read(12, "\231option \302\253\302\240-o\302\240\302\273.\\n\"\nmsgid \"%s"..., 4096) = 4096
> read(12, " ignored!\\n\"\nmsgstr \"%s: Attenti"..., 4096) = 4096
> read(12, "mat \302\253 %s \302\273.\"\nmsgid \"Cover open"..., 4096) = 4096
> read(12, "\"ERROR: l\342\200\231h\303\264te \303\240 distance n\342"..., 4096) = 4096
> read(12, "R: Unknown format character \\\"%c"..., 4096) = 4096
> read(12, " d\342\200\231impression...\\n\"\nmsgid \"INF"..., 4096) = 4096
> read(12, " )\"\nmsgid \"Job #%d cannot be res"..., 4096) = 4096
> read(12, " pointer\"\nmsgstr \"Pointeur de fi"..., 4096) = 4096
> read(12, "\nmsgstr \"D\303\251finir les autorisati"..., 4096) = 4096
> read(12, "ble to copy Windows 2000 printer"..., 4096) = 4096
> read(12, "           [-r remove-class] [-v"..., 4096) = 4096
> read(12, " [ options ] nom de fichier(s)\\n"..., 4096) = 4096
> read(12, "a pas r\303\251pondu avec un octet d\342\200"..., 4096) = 4096
> read(12, "\302\240\302\273.\"\nmsgid \"lpadmin: Class nam"..., 4096) = 4096
> read(12, "to set the device URI:\\n        "..., 4096) = 4096
> brk(0xc2000)                            = 0xc2000
> read(12, "gid \"lppasswd: Unable to open pa"..., 4096) = 3186
> read(12, "", 4096)                      = 0
> close(12)                               = 0
> time(NULL)                              = 1246110922
> write(5, "D [27/Jun/2009:15:55:22 +0200] c"..., 80) = 80
> stat64("/opt/share/doc/cups/fr/images/top-left.gif", 0xbefefac8) = -1 ENOENT (No such file or directory)
> stat64("/opt/share/doc/cups/fr/images/top-left.gif", 0xbefefac8) = -1 ENOENT (No such file or directory)
> stat64("/opt/share/doc/cups/images/top-left.gif", {st_mode=S_IFREG|0644, st_size=17592186044416, ...}) = 0
> open("/opt/share/doc/cups/images/top-left.gif", O_RDONLY|O_LARGEFILE) = 12
> fcntl64(12, F_GETFD)                    = 0
> fcntl64(12, F_SETFD, FD_CLOEXEC)        = 0
> read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
> read(12, "", 4096)                      = 0
> _llseek(12, 0, [0], SEEK_SET)           = 0
> read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
> read(12, "", 4096)                      = 0
> _llseek(12, 0, [0], SEEK_SET)           = 0
> read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
> read(12, "", 4096)                      = 0
> _llseek(12, 0, [0], SEEK_SET)           = 0
> read(12, "GIF87a\17\0P\0\245\"\0\0\0\0\f\f\t\31\31\22%%\03322$>>-K"..., 4096) = 473
> read(12, "", 4096)                      = 0
> close(12)                               = 0
> open("/opt/share/doc/cups/images/top-left.gif", O_RDONLY|O_LARGEFILE) = 12
> fcntl64(12, F_GETFD)                    = 0
> fcntl64(12, F_SETFD, FD_CLOEXEC)        = 0
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> time(NULL)                              = 1246110922
> send(11, "HTTP/1.1 200 OK\r\nDate: Sat, 27 J"..., 236, 0) = 236
> epoll_ctl(0x3, 0x3, 0xb, 0xbefefb50)    = 0
> --- SIGSEGV (Segmentation fault) @ 0 (0) ---
> +++ killed by SIGSEGV +++
>

[phild_mande]

Hi Brian,
--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> It is known from users of transmission, that a few platforms using cs05q1armel feed need to set EVENT_NOEPOLL with a non-empty value, to work around the (possibly kernel) problem.
>
> But since cups does not use libevent, for this feed, probably better to disable epoll during configure with ac_cv_func_epoll_create=no (before line 189 of make/cups.mk)?  I don't have a device to test, someone should give it a try.
>
> Good luck,
>
> -Brian
>
Unfortunatly, at this time, I have only a cs08q1armel env. and for trying compilation under crosscompile codesourcery for piece of software outside ipkg ... so not sure that I am able to do the job ...
BUT if anybody can make a compile of cups, with this modification .. I can test when you want ... I have a TS-109 available for BETA version and to do all test I want.

In parallel I try to set up a ipkg compile env. ... but not sure of the delay

In all case Thanks a lot for your help and time spend to help user like me.

Philippe.

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> But since cups does not use libevent, for this feed, probably better to disable epoll during configure with ac_cv_func_epoll_create=no (before line 189 of make/cups.mk)?  I don't have a device to test, someone should give it a try.
>
> Good luck,
>
> -Brian
>
Hi Brian,

I have installed svn and tools to build cups ... after some small error (from me), I have started to compile cups under cs05q1armel
I have introduce the ac_cv_func_epoll_create=no just before line :
./configure \ ... I add a \ as I suppose

all seems running except I stop on an error when we get the jpegsrc.v6b.tar.gz, because the file is not found ... searching in www.ijg.org ... seems that now it's v7 ...
Please what I have to change to continue ....

Thanks for your help.

Philippe.

[phild_mande]

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:

>
> --- In nslu2-linux@..., "Brian Zhou" <b88zhou@> wrote:
> >
> > But since cups does not use libevent, for this feed, probably better to disable epoll during configure with ac_cv_func_epoll_create=no (before line 189 of make/cups.mk)?  I don't have a device to test, someone should give it a try.
> >
> > Good luck,
> >
> > -Brian
> >
> Hi Brian,
>
> I have installed svn and tools to build cups ... after some small error (from me), I have started to compile cups under cs05q1armel
> I have introduce the ac_cv_func_epoll_create=no just before line :
> ./configure \ ... I add a \ as I suppose
>
> all seems running except I stop on an error when we get the jpegsrc.v6b.tar.gz, because the file is not found ... searching in www.ijg.org ... seems that now it's v7 ...
> Please what I have to change to continue ....
>
> Thanks for your help.
>
> Philippe.
>

OK, I have found ... changing version number in libjpeg.mk

Sorry for the question, it's first time to me to used ipkg compile tools.

Philippe.

[phild_mande]

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:

>
> --- In nslu2-linux@..., "phild_mande" <father_mande@> wrote:
> >
> > --- In nslu2-linux@..., "Brian Zhou" <b88zhou@> wrote:
> > >
> > > But since cups does not use libevent, for this feed, probably better to disable epoll during configure with ac_cv_func_epoll_create=no (before line 189 of make/cups.mk)?  I don't have a device to test, someone should give it a try.
> > >
> > > Good luck,
> > >
> > > -Brian

> > I have installed svn and tools to build cups ... after some small error (from me), I have started to compile cups under cs05q1armel
> > I have introduce the ac_cv_func_epoll_create=no just before line :
> > ./configure \ ... I add a \ as I suppose
> >
> > all seems running except I stop on an error when we get the jpegsrc.v6b.tar.gz, because the file is not found ... searching in www.ijg.org ... seems that now it's v7 ...
> > Please what I have to change to continue ....
> >
> >
> OK, I have found ... changing version number in libjpeg.mk
>
> Sorry for the question, it's first time to me to used ipkg compile tools.
>
> Philippe.
>

SUCCEDD ... Whaou... Brian ... you are the best ....
after compiling with this option, upgrading libjpeg to v7 to be coherent ...
I have created the ipk
install libjpeg V7 ... normal it's an upgrade ...
install with -force-reinstall cups with the modif ... seg fault using web interface desappear ...

I am able now to configure my printer using cups interface ... without problems ... even if I only do partial test ...

VERY GOOD WORK ...

Now what we have to do to install this modification ???? in feeds

Philippe.

[brian_zhou]

--- In nslu2-linux@..., "phild_mande" <father_mande@...> wrote:
> SUCCEDD ... Whaou... Brian ... you are the best ....
> after compiling with this option, upgrading libjpeg to v7 to be coherent ...
> I have created the ipk
> install libjpeg V7 ... normal it's an upgrade ...
> install with -force-reinstall cups with the modif ... seg fault using web interface desappear ...
>
> I am able now to configure my printer using cups interface ... without problems ... even if I only do partial test ...
>

Excellent progress! You are now qualified as an optware developer. I can give you SVN commit access if you'd like.

>
> Now what we have to do to install this modification ???? in feeds
>
Just made two commits

  http://trac.nslu2-linux.org/optware/changeset/10395
  http://trac.nslu2-linux.org/optware/changeset/10396

to integrate your change. Please note that I've not upgraded libjpeg to 7 just yet, since it will affect quite a few packages. I plan to do that a little later.

Best regards,

-Brian

[phild_mande]

--- In nslu2-linux@..., "Brian Zhou" <b88zhou@...> wrote:
>
> Excellent progress! You are now qualified as an optware developer. I can give you SVN commit access if you'd like.
>

Appreciate, but not sure that it's true, lot of diff. between doing modification suggest by expert ... and doing all the job ... but in all case I continue to test and, even I try to create my own ipk ... small prog like vncrepeater, that I have yet compile for TS-x09, just to see what append and if I understand what we have to do ... for ex. I just know now that they have a save of previous source ... so no need to update for nothing libjpeg ...

> >
> > Now what we have to do to install this modification ???? in feeds
> >
> Just made two commits
>
>   http://trac.nslu2-linux.org/optware/changeset/10395
>   http://trac.nslu2-linux.org/optware/changeset/10396
>
> to integrate your change. Please note that I've not upgraded libjpeg to 7 just yet, since it will affect quite a few packages. I plan to do that a little later.
>
> Best regards,
>
> -Brian
>

Thanks again for  the modif ... it's, for me, an "online student course" about ipk ... nice.

If somebody have problem on TS-x09 or TS-x19 and you want to do some test ... feelfree to send me a mail, if I have the possibility, I can run test on my device for you.

Regards.
Philippe.