hpijs crashed with SIGSEGV in memcpy() - after using two sided printing [Jaunty]

[unggnu]

Binary package hint: hplip

Hpijs crashes always if I print a 2 page pdf on one paper with two sided long-edge. If I use normal one sided printing it doesn't crash.

I am using a HP Photosmart 3300 with a two sided printing module.

The document was http://www.drivecity.de/media/content/drivecity_rma.pdf .

My default printer format is not letter. I don't know where apport gets this value.

ProblemType: Crash
Architecture: amd64
DistroRelease: Ubuntu 9.04
ExecutablePath: /usr/bin/hpijs
Lpstat:
 device for Photosmart_3300: hp:/net/Photosmart_3300_series?ip=10.0.0.251
 device for Photosmart_3300_fax: hpfax:/net/Photosmart_3300_series?ip=10.0.0.251
MachineType: Gigabyte Technology Co., Ltd. EP45-DS3L
Package: hpijs 3.9.2-3ubuntu3
Papersize: letter
PpdFiles:
 Photosmart_3300_fax: HP Fax
 Photosmart_3300: HP Photosmart 3300 Series hpijs, hpijs 3.9.2
ProcCmdLine: root=/dev/mapper/root ro quiet splash
ProcCmdline: hpijs
ProcEnviron:
 PATH=(custom, no user)
 LANG=en_US.UTF8
ProcVersionSignature: Ubuntu 2.6.28-11.39-generic
Signal: 11
SourcePackage: hplip
StacktraceTop:
 memcpy () at ../sysdeps/x86_64/memcpy.S:267
 UXServices::CopyData (this=0x19fd920,
 UXServices::ToDevice (this=0x19fd920,
 apdk::Printer::Send (this=0x1a053c0,
 apdk::Printer::Encapsulate (this=0x1a053c0,
Title: hpijs crashed with SIGSEGV in memcpy()
UserGroups:

[unggnu (unggnu)]

Since I already installed the libc6 and hplip dbg packages I post the ThreadStacktrace here if the apport service removes it after retracing.

.
Thread 1 (process 19685):
#0 memcpy () at ../sysdeps/x86_64/memcpy.S:267
No locals.
#1 0x0000000000439c7a in UXServices::CopyData (this=0x19fd920,
    pBuffer=0x1a91680 "�\026", iCount=4096) at /usr/include/bits/string3.h:52
 p = <value optimized out>
#2 0x000000000043a1de in UXServices::ToDevice (this=0x19fd920,
    pBuffer=0x1a91680 "�\026", Count=0x7fff8f50eb18)
    at prnt/hpijs/services.cpp:342
 cnt = 0
#3 0x0000000000417836 in apdk::Printer::Send (this=0x1a053c0,
    pWriteBuff=0x1a9de80 "\237�\026", dwWriteCount=4)
    at prnt/hpijs/printer.cpp:668
 write_error = apdk::NO_ERROR
 residual = 4096
 status_reg = 0 '\0'
 eDisplayStatus = apdk::DISPLAY_PRINTING
 JobCanceled = 0
 pBuffer = (const BYTE *) 0x1a91680 "�\026"
 dwSendSize = 4096
 BytesToWrite = 3
#4 0x0000000000417fc3 in apdk::Printer::Encapsulate (this=0x1a053c0,
    InputRaster=0x1a7a6d0, bLastPlane=<value optimized out>)
    at prnt/hpijs/printer.cpp:410
 err = <value optimized out>
 scratch = "\033*b4V\000\000\000\001\000\000\000\000\000\000\000Ц�\001"
 scratchLen = <value optimized out>
#5 0x000000000041c2d8 in apdk::RasterSender::Process (this=0x1a7a6b0,
    InputRaster=0x1a91680) at prnt/hpijs/translator.cpp:56
 err = <value optimized out>
 bOutput = <value optimized out>
#6 0x0000000000415185 in apdk::Pipeline::Execute (this=0x0,
    InputRaster=0x1a91680) at prnt/hpijs/job.cpp:1026
No locals.
#7 0x00000000004151fa in apdk::Pipeline::Execute (this=0x1a76a70,
    InputRaster=<value optimized out>) at prnt/hpijs/job.cpp:1038
No locals.
#8 0x00000000004151fa in apdk::Pipeline::Execute (this=0x1a7b850,
    InputRaster=<value optimized out>) at prnt/hpijs/job.cpp:1038
No locals.
#9 0x00000000004151fa in apdk::Pipeline::Execute (this=0x1a8a540,
    InputRaster=<value optimized out>) at prnt/hpijs/job.cpp:1038
No locals.
#10 0x000000000043a6f2 in UXServices::SendBackPage (this=0x19fd920)
    at prnt/hpijs/services.cpp:75
 err = <value optimized out>
 i = 83
#11 0x0000000000435f74 in main (argc=<value optimized out>,
    argv=<value optimized out>, evenp=<value optimized out>)
    at prnt/hpijs/hpijs.cpp:753
 pSS = (class UXServices *) 0x19fd920
 ctx = (IjsServerCtx *) 0x19fb850
 raster = 0xbb8 <Address 0xbb8 out of bounds>
 status = <value optimized out>
 ret = <value optimized out>
 n = 0
 i = 3000
 kn = 0
 width = <value optimized out>
 k_width = 310
 pDev = <value optimized out>

[Till Kamppeter (till-kamppeter)]

Can someone at HP look into this? Looks like that duplex printing on inkjets does not work.

[Apport retracing service (apport)]

StacktraceTop:memcpy () at ../sysdeps/x86_64/memcpy.S:267
UXServices::CopyData (this=0x19fd920,
UXServices::ToDevice (this=0x19fd920,
apdk::Printer::Send (this=0x1a053c0,
apdk::Printer::Encapsulate (this=0x1a053c0,

[Jason Callough (jcallough)]

We have been unable to reproduce this issue up to this point but believe that we have an understanding as to why the crash is occuring. Please check in periodically as we are currently working on a patch to correct the problem.

Jason

[unggnu (unggnu)]

Thanks for trying to nail down the issue.

Btw. the first page gets printed and it seems to crash while the five second dry time or shortly afterwards. After this the paper is collected again and pushed out without printing the second page.

[Jason Callough (jcallough)]

We have been unable to reproduce this but we are aware that there is an issue regarding duplex printing in Jaunty. Therefore, keep an eye out for the next release coming soon where this will be addressed.

Jason

[Jason Callough (jcallough)]

Check for the next release and update accordingly.

Jason

[Till Kamppeter (till-kamppeter)]

Can you post a patch for HPIJS so that we can do a stable release update (SRU) for Jaunty as soon as Jaunty gets released? Thanks in advance.