Loading ...
Sorry, an error occurred while loading the content.
 

Re: Can't add cscope connection

Expand Messages
  • David Rennalls
    ... I only had patches 1-214, I just built a version with patches 1-223 and tried. I got the same result however. - David
    Message 1 of 11 , Feb 2, 2004
      Bram Moolenaar wrote:

      > David Rennalls wrote:
      >
      >
      >>Khorev Sergey wrote:
      >>
      >>
      >>>> When doing a ':cs add cscope.out' where cscope.out is
      >>>>in the current directory all I get is the following error.
      >>>>E262: error reading cscope connection 0
      >>>>Runing cscope separately works fine however. Any ideas ?
      >>>
      >>>
      >>>What says "cscope -dl -f cscope.out"?
      >>>
      >>
      >>I get the '>>' prompt for cscope's line-oriented interface. Doing a
      >>'1MySymbol' correctly shows me occurrences of 'MySymbol'.
      >
      >
      > Did you try with or without patch 6.2.223? Then try without/with it.
      >

      I only had patches 1-214, I just built a version with patches 1-223 and
      tried. I got the same result however.

      - David
    • Khorev Sergey
      Which of the cscopes flying around are you using? Operating system? Can you make repeatable test case? If not, use debugger to reveal stack trace on the call
      Message 2 of 11 , Feb 2, 2004
        Which of the cscopes flying around are you using?
        Operating system?
        Can you make repeatable test case?
        If not, use debugger to reveal stack trace on the call to cs_reading_emsg()

        --
        Sergey Khorev
        http://iamphet.nm.ru
        ROS: Return On Shield
        -- New Opcode
      • David Rennalls
        ... I did some quick stepping to find out where exactly it s failing. Basically it s in cs_read_prompt, in if_cscope.c line 1989. There s a while loop that
        Message 3 of 11 , Feb 6, 2004
          Khorev Sergey wrote:
          > Which of the cscopes flying around are you using?
          > Operating system?
          > Can you make repeatable test case?
          > If not, use debugger to reveal stack trace on the call to cs_reading_emsg()
          >
          I did some quick stepping to find out where exactly it's failing.
          Basically it's in cs_read_prompt, in if_cscope.c line 1989. There's a
          while loop that seems to be used to skip over the the ">> "
          cscope prompt.
          ....
          while ((ch = getc(csinfo[i].fr_fp)) != EOF && ch != CSCOPE_PROMPT[0])
          ....

          But it's getting EOF from the getc on the first iteration. So further
          down the check on line 2024 ...if (ch == EOF)... is failing. After the
          failing call to getc csinfo[i].fr_fp._flag is 0x19. I don't have time to
          debug further right now but hopefully that helps.
          The debug version I built is 6.2.238.

          cs_read_prompt(int 0x00000000) line 1989
          cs_add_common(char * 0x010de8c7, char * 0x00000000, char * 0x00000000)
          line 519 + 26 bytes
          cs_add(exarg * 0x0012f880) line 397 + 17 bytes
          do_cscope_general(exarg * 0x0012f880, int 0x00000000) line 131 + 10 bytes
          do_cscope(exarg * 0x0012f880) line 141 + 11 bytes
          do_one_cmd(unsigned char * * 0x0012f9b4, int 0x00000000, condstack *
          0x0012f9c0, unsigned char * (int, void *, int)* 0x004465b8
          getexline(int, void *, int), void * 0x00000000) line 2444 + 16 bytes
          do_cmdline(unsigned char * 0x00000000, unsigned char * (int, void *,
          int)* 0x004465b8 getexline(int, void *, int), void * 0x00000000, int
          0x00000000) line 977 + 40 bytes
          nv_colon(cmdarg * 0x0012fc54) line 4629 + 29 bytes
          normal_cmd(oparg * 0x0012fcb4, int 0x00000001) line 1108 + 16 bytes
          main_loop(int 0x00000000) line 2135 + 11 bytes
          VimMain() line 1953 + 7 bytes
          WinMain(HINSTANCE__ * 0x00400000, HINSTANCE__ * 0x00000000, char *
          0x00132f7f, int 0x00000001) line 125 + 14 bytes
          WinMainCRTStartup() line 198 + 54 bytes
          KERNEL32! 7c5987e7()
          z
          - David
        • Gary Johnson
          ... The :version output from your original post indicated that you were using Windows, and the stack trace below indicates that that s still the case. ...
          Message 4 of 11 , Feb 6, 2004
            On 2004-02-06, David Rennalls <drennalls@...> wrote:
            > Khorev Sergey wrote:
            > >Which of the cscopes flying around are you using?
            > >Operating system?

            The :version output from your original post indicated that you were
            using Windows, and the stack trace below indicates that that's still
            the case.

            > >Can you make repeatable test case?
            > >If not, use debugger to reveal stack trace on the call to cs_reading_emsg()
            > >
            > I did some quick stepping to find out where exactly it's failing.
            > Basically it's in cs_read_prompt, in if_cscope.c line 1989. There's a
            > while loop that seems to be used to skip over the the ">> "
            > cscope prompt.
            > ....
            > while ((ch = getc(csinfo[i].fr_fp)) != EOF && ch != CSCOPE_PROMPT[0])
            > ....
            >
            > But it's getting EOF from the getc on the first iteration. So further
            > down the check on line 2024 ...if (ch == EOF)... is failing. After the
            > failing call to getc csinfo[i].fr_fp._flag is 0x19. I don't have time to
            > debug further right now but hopefully that helps.
            > The debug version I built is 6.2.238.
            >
            > cs_read_prompt(int 0x00000000) line 1989
            > cs_add_common(char * 0x010de8c7, char * 0x00000000, char * 0x00000000)
            > line 519 + 26 bytes
            > cs_add(exarg * 0x0012f880) line 397 + 17 bytes
            > do_cscope_general(exarg * 0x0012f880, int 0x00000000) line 131 + 10 bytes
            > do_cscope(exarg * 0x0012f880) line 141 + 11 bytes
            > do_one_cmd(unsigned char * * 0x0012f9b4, int 0x00000000, condstack *
            > 0x0012f9c0, unsigned char * (int, void *, int)* 0x004465b8
            > getexline(int, void *, int), void * 0x00000000) line 2444 + 16 bytes
            > do_cmdline(unsigned char * 0x00000000, unsigned char * (int, void *,
            > int)* 0x004465b8 getexline(int, void *, int), void * 0x00000000, int
            > 0x00000000) line 977 + 40 bytes
            > nv_colon(cmdarg * 0x0012fc54) line 4629 + 29 bytes
            > normal_cmd(oparg * 0x0012fcb4, int 0x00000001) line 1108 + 16 bytes
            > main_loop(int 0x00000000) line 2135 + 11 bytes
            > VimMain() line 1953 + 7 bytes
            > WinMain(HINSTANCE__ * 0x00400000, HINSTANCE__ * 0x00000000, char *
            > 0x00132f7f, int 0x00000001) line 125 + 14 bytes
            > WinMainCRTStartup() line 198 + 54 bytes
            > KERNEL32! 7c5987e7()
            > z

            That's where I thought it was failing--it's good to have this
            confirmation. cs_add_common() calls cs_read_prompt() right after it
            calls cs_create_connection() at line 518 of if_cscope.c. It seems
            to me that the likely thing that's failing is the call to spawnlp()
            at either line 875 or 877. I say that because I think that's the
            part of cs_add_common() that's most sensitive to your environment.
            If the problem was in the cs_add_common() code itself, I would
            expect Sergey or another Windows user to have had problems before
            now. However, I don't have cscope on my Windows machine and I know
            zilch about Windows task management or interprocess communications,
            so I don't know how to figure out why launching cscope from
            spawnlp() seems to fail while launching it from the command line
            works. I'm sure Sergey can provide a lot more insight here.

            HTH,
            Gary

            --
            Gary Johnson | Agilent Technologies
            garyjohn@... | Wireless Division
            | Spokane, Washington, USA
          Your message has been successfully submitted and would be delivered to recipients shortly.