artifact_id;status_id;status_name;priority;submitter_id;submitter_name;assigned_to_id;assigned_to_name;open_date;close_date;last_modified_date;summary;details;"Category";"Group";"Resolution" 301211;1;"Open";3;2380;"Michael Banck";100;"Nobody";"2005-02-20 12:46";"";"2008-04-17 14:40";"#296115: ifupdown (debian-ports)";"Author: Michael Banck Source: ifupdown Status: submitted(#296115) Categories: debian Strip-Level: -p1";"debian";"submitted";"None" 301248;1;"Open";3;3630;"Marc Dequènes";100;"Nobody";"2005-03-10 00:49";"";"2006-10-27 15:35";"[HurdFr] cdebootstrap FTBFS";"Author: Marc Dequènes (Duck) Source: cdebootstrap Status: unsubmitted Categories: other, needs-more-testing Strip-Level: -p1 PATH_MAX & include fixes. Needs a configure check to add missing include instead of ""#ifndef __GNU__"". Depends on hurd-select problem in glibc being solved. ";"other";"unsubmitted";"None" 301321;1;"Open";3;11528;"Christopher Bodenstein";100;"Nobody";"2005-03-28 13:52";"";"2006-10-27 15:35";"FTBFS gkrellm";"The following patch allows gkrellm to build on the Hurd. The patch to the top-level Makefile is to detect and use libgtop-1.0 (which isn't found otherwise). The rest of the patch is to get HAVE_GETADDRINFO defined. The original code uses #if defined(__linux__), so I've just added a #if defined(__GNU__) which is probably not good but works. I haven't had the opportunity yet to check if the package actually works, but it's available from http://packages.physicman.net/ for testing.";"linuxism";"unreviewed";"None" 302870;1;"Open";3;2380;"Michael Banck";100;"Nobody";"2006-01-08 14:05";"";"2006-10-27 15:35";"#346447: gopher";"Author: Ben Asselstine Source: gopher Status: Submitted (#346447) Maturity: Unreviewed Strip-Level: -p1 Hurd-Specific: yes";"posix";"submitted";"None" 303152;1;"Open";3;11528;"Christopher Bodenstein";100;"Nobody";"2006-03-01 15:31";"";"2006-10-27 15:35";"cdargs";"Author: Christopher Bodenstein Source: cdargs Status: unreviewed Categories: posix Strip-Level: -p1 The attached patch uses get_current_dir_name() instead of getcwd if _GNU_SOURCE is defined (get_current_dir_name should be tested in the configure but I can't get it to work properly, when I modify the configure.in the resulting configure is all messed up and the package won't build) and avoids the use of PATH_MAX in that case. ";"posix";"unreviewed";"None" 303178;1;"Open";3;11521;"Barry deFreese ";100;"Nobody";"2006-03-05 21:50";"";"2006-10-27 15:35";"jamvm-1.4.2";"Author: Barry deFreese Source: jamvm Status: unreviewed Categories: build-env Strip-Level: -p1 Missing configure options for gnu*)";"build-env";"unreviewed";"None" 303210;1;"Open";3;11528;"Christopher Bodenstein";100;"Nobody";"2006-03-09 09:17";"";"2006-10-27 15:35";"mailutils";"Author: Christopher Bodenstein Source: mailutils Strip-Level: -p1 Status: unreviewed Categories: posix The attached patch removes one instance of MAXHOSTNAMELEN using xgethostname(). Apparently, you need automake1.9 to get the .am files to be processed correctly. Also, I had to disable the checks in the debian/rules file as at least one of them seems to make my machine freeze. ";"posix";"unreviewed";"None" 303220;1;"Open";3;11528;"Christopher Bodenstein";100;"Nobody";"2006-03-10 16:24";"";"2006-10-27 15:35";"dvdisaster";"Author: Christopher Bodenstein Source: dvdisaster Strip-Level: -p1 Status: unreviewed Categories: posix There is one instance of PATH_MAX used in a getcwd(). I use a get_current_dir_name() to replace it, however there is no configure.in so I patched configure directly to check the the function's presence.";"posix";"unreviewed";"None" 303255;1;"Open";3;11521;"Barry deFreese ";100;"Nobody";"2006-03-19 19:09";"";"2006-10-27 15:35";"libusb-0.1.11";"Here is a patch to build libusb on GNU/Hurd. Obviously you need to run aclocal, autoheader, and autoreconf. It isn't quite complete becuase it didn't finish building but I was getting a segfault with jade making some of the files.";"several";"unreviewed";"None" 303733;1;"Open";3;2380;"Michael Banck";100;"Nobody";"2006-08-08 21:20";"";"2006-10-27 15:35";"cdargs";"Author: Ben Asselstine ";"posix";"unreviewed";"None" 303738;1;"Open";3;2380;"Michael Banck";100;"Nobody";"2006-08-08 21:26";"";"2006-10-27 15:35";"postal";"Author: Ben Asselstine the postal folks have gnu_source defined in other places, so this is okay i think (to add it in logit.cpp) ";"posix";"unreviewed";"None" 304336;1;"Open";3;11521;"Barry deFreese ";100;"Nobody";"2007-04-12 17:27";"";"2007-04-12 17:27";"[FTBFS] ntp 4.2.2.p4+dfsg";"Author: Barry deFreese Source: ntp Status: Submitting to BTS Categories: build-env Strip-Level: -p1 Had to disable IPv6 support and linuxcap in ./configure. Also it was trying to move ntptime binary which we can't build because of missing timex.h from our glibc. ";"build-env";"submitted";"Fixed" 305130;1;"Open";3;12824;"Samuel Thibault";100;"Nobody";"2007-09-29 13:20";"2007-12-22 22:20";"2009-11-17 00:51";"native backend for hal";"Fixes compilation, but still using the dummy backend. We need to port hal on the Hurd, but we should be probably don't do it blindly, build nice hurdish abstractions, etc. ";"None";"None";"Later" 310749;1;"Open";3;201950;"Flávio Cruz";100;"Nobody";"2008-04-30 23:18";"";"2008-05-06 00:13";"#479657 dancer-ircd patch";"dancer-ircd-1.0.36 patch for GNU/Hurd";"None";"None";"None" 311309;1;"Open";3;12824;"Samuel Thibault";100;"Nobody";"2008-12-16 02:01";"";"2009-01-17 03:06";"509252: ghc6 FTBFS (debian-ports)";"This is due to ghc using setitimer(TIMER_VIRTUAL), for which Mach doesn't provide any support. The attached patch just makes it use ITIMER_REAL instead. I've filed a ticket upstream http://hackage.haskell.org/trac/ghc/ticket/2883 Maybe this could be a debian-ports candidate, as ghc6 is blocking all haskell packages. ";"None";"None";"None" 311742;1;"Open";3;204851;"Barry deFreese";100;"Nobody";"2009-06-09 14:59";"";"2009-06-09 14:59";"FTBFS scheme2c 1993.3.15.2-4";"By using the following change to debian/rules I can get it to build further: arch := $(shell dpkg-architecture -qDEB_BUILD_ARCH) ifeq ($(arch),amd64) S2C_ARCH = AMD64 else ifeq ($(arch),i386) S2C_ARCH = LINUX +else ifeq ($(arch),hurd-i386) + S2C_ARCH = LINUX else $(warning Warning: unsupported architecture $(arch)) S2C_ARCH = UNSUPPORTED endif However, it fails with the following: gcc -c -m32 -O2 -march=i386 -I. cio.c cio.c:87:2: error: #error ""Stack overflow is enabled for this architecture but you are either missing libsigsegv or do not have stack overflow and sigsegv recovery built in"" cio.c: In function 'sc_segv__handlers': cio.c:842: error: 'stackoverflow_handler' undeclared (first use in this function) cio.c:842: error: (Each undeclared identifier is reported only once cio.c:842: error: for each function it appears in.) cio.c:843: error: 'overflow_stack' undeclared (first use in this function) cio.c:846: error: 'sigsegv_handler' undeclared (first use in this function) make[3]: *** [cio.o] Error 1 make[3]: Leaving directory `/devel/bdefreese/scheme2c/scheme2c-1993.3.15.2/LINUX/scrt' make[2]: *** [port] Error 2 make[2]: Leaving directory `/devel/bdefreese/scheme2c/scheme2c-1993.3.15.2/LINUX/scrt' make[1]: *** [port] Error 2 make[1]: Leaving directory `/devel/bdefreese/scheme2c/scheme2c-1993.3.15.2/LINUX' make: *** [build-arch-stamp] Error 2 dpkg-buildpackage: error: debian/rules build gave error exit status 2 Barry";"linuxism";"unreviewed";"None" 311743;1;"Open";3;204851;"Barry deFreese";100;"Nobody";"2009-06-09 15:51";"";"2009-06-09 15:51";"FTBFS partimage 0.6.7-2";"Using the attached patch, I get it past the unsupported OS. However, it still uses MAXPATHLEN. ";"linuxism";"unreviewed";"None" 311759;1;"Open";3;204851;"Barry deFreese";204851;"Barry deFreese";"2009-06-17 20:32";"";"2009-06-20 02:32";"[FTBFS] #533731 hpsockd 0.17";"Getting hpsockd to recognize the system is as easy as adding the following to src/Makefile.os: GNU: Linux GNU.debug: Linux.debug However, there is then an unconditional use of MAXPATHLEN. src/sockd/gram.y: hostToken: T_HOST { struct hostent *hp; char tmp[MAXPATHLEN]; register int i; gethostname(tmp,sizeof(tmp)); if ((hp=gethostbyname(tmp))==NULL) { syslog(LOG_ERR,""No such host""); YYABORT; } NEWLIST($$,*(int*)hp->h_addr_list[0]); for (i=1; hp->h_addr_list[i]; i++) ADDTOLIST($$,*(int*)hp->h_addr_list[i]); } Why it uses MAXPATHLEN here is beyond me. Barry";"linuxism";"submitted";"None" 311761;1;"Open";3;204851;"Barry deFreese";100;"Nobody";"2009-06-19 18:24";"";"2012-06-01 01:57";"Partial patch for kbuild 1:0.1.98svn2318-1 ";"Hi, Attached is the start of a patch for kbuild. Unfortunately it has several PATH_MAX issues that I just #ifdef'd for now. But even larger is that it include sys/mount.h in several places. Thanks, Barry";"linuxism";"submitted";"None" 311762;1;"Open";3;204851;"Barry deFreese";100;"Nobody";"2009-06-19 19:27";"";"2009-06-19 19:27";"Patch for libace-perl 1.92-1 (Still has linuxisms)";"Hi, Attached is a patch for libace-perl. Currently I still have #ifndefs in for the MAXPATHLEN issues so those need to be fixed yet. Thanks, Barry ";"linuxism";"unreviewed";"None" 311767;1;"Open";3;204851;"Barry deFreese";100;"Nobody";"2009-06-21 04:55";"";"2009-06-21 15:18";"[FTBFS] #533915 nrss 0.3.9-1";"Does this patch for nrss look reasonable? Thanks, Barry deFreese";"build-env";"submitted";"None" 312117;1;"Open";3;202476;"Pino Toscano";100;"Nobody";"2009-11-29 11:32";"";"2009-11-29 13:32";"#558586: hyperestraier 1.4.9";"Simple malloc+free usage instead of hardcoded PATH_MAX-sized strings.";"None";"None";"None" 312442;1;"Open";3;210375;"David Hedberg";100;"Nobody";"2010-04-06 19:53";"";"2011-04-09 13:47";"libcap2 2.19";"This small patch allows libcap to build and be linked with the accompanying utilities in progs. I have not tested it with anything else and there is of course no real functionality provided.";"None";"None";"None" 312488;1;"Open";3;208675;"Gokturk Yuksek";100;"Nobody";"2010-05-04 08:54";"";"2010-05-04 08:54";"sound/wsynth-dssi-0.1.3";"Proposed patch solves the PATH_MAX issue by determining the size of the buffer using 'strlen()' first and then allocating the buffer dynamically.";"posix";"None";"None"