Re: compilation hangs on Ubuntu

[Eugène Adell <eugene.adell () gmail com>]

Submitted and validated by Guy Harris. All's fine now.

2018-05-14 20:29 GMT+02:00 Dario Lombardo <lomato () gmail com>:

> Can you submit a change on Gerrit for it?
>
> On Mon, May 14, 2018, 19:58 Eugène Adell <eugene.adell () gmail com> wrote:
>
> > Thanks all for the help and Darius for finding out this conflict with
> > ninja. Building with an ordinary user works.
> >
> >
> > By the way, I have another small request. Could you please change the
> > AUTHORS file to keep my personal address (my email at D2-SI didn't survive
> > after I left, and I contributed twice, the first time with D2-SI and later
> > under my personal address). Thanks.
> >
> > kind regards
> > Eugène
> >
> >
> >
> >
> > 2018-05-14 17:11 GMT+02:00 Darius Davis <darius () vmware com>:
> >
> > > Hi Eugène,
> > >
> > >
> > >
> > > The strace log shows that your Ubuntu 16.04 system has the "ninja"
> > > package installed, which is "a privilege escalation detection and
> > > prevention system for GNU/Linux hosts" (https://packages.ubuntu.com/
> > > xenial/admin/ninja).  It installs into /usr/sbin/ninja.  Your build
> > > problem is caused by the presence of that "ninja" package on your system --
> > > and the fact that you appear to be building wireshark as "root".
> > >
> > >
> > >
> > > While running as "root", your PATH contains /usr/sbin, and that entry
> > > appears before /usr/bin, so instead of cmake launching the ninja build tool
> > > (/usr/bin/ninja), it launches the ninja privilege escalation detection tool
> > > (/usr/sbin/ninja), which busily monitors your /proc directory for evidence
> > > of system intrusion... but does not build Wireshark.
> > >
> > >
> > >
> > > 1900  access("/usr/sbin/ninja", R_OK)   = 0
> > >
> > > 1900  stat("/usr/sbin/ninja", {st_mode=S_IFREG|0755, st_size=22872,
> > > ...}) = 0
> > >
> > > [...]
> > >
> > > 1901  execve("/usr/sbin/ninja", ["/usr/sbin/ninja", "--version"], [/* 22
> > > vars */] <unfinished ...>
> > >
> > >
> > >
> > > You should be able to resolve the issue by building Wireshark as a
> > > non-root user -- I don't think there is any need or reason to build as
> > > "root".  As a regular user, your PATH should not contain /usr/sbin, and the
> > > ninja intrusion-detection program will be ignored.  cmake should then find
> > > and launch the ninja build tool correctly.
> > >
> > >
> > >
> > > If you are not going to use the "ninja" privilege escalation detection
> > > program (for example, if it was installed by accident while you were
> > > looking for the "ninja-build" package), you can uninstall it by running
> > > "sudo apt-get remove ninja", and that would resolve the build problem too.
> > > (Regardless, I would really advise not building as "root" unless you have a
> > > very good reason to do so!)
> > >
> > >
> > >
> > > Cheers,
> > >
> > > --
> > >
> > > Darius
> > >
> > >
> > >
> > > *From: *Wireshark-dev <wireshark-dev-bounces () wireshark org> on behalf
> > > of Eugène Adell <eugene.adell () gmail com>
> > > *Reply-To: *Developer support list for Wireshark <
> > > wireshark-dev () wireshark org>
> > > *Date: *Tuesday, 15 May 2018 at 12:25 am
> > > *To: *Developer support list for Wireshark <wireshark-dev () wireshark org>
> > > *Subject: *Re: [Wireshark-dev] compilation hangs on Ubuntu
> > >
> > >
> > >
> > > It doesn't hang, it returns :
> > >
> > > 2;7;12
> > >
> > > As requested, the strace is attached.
> > >
> > >
> > >
> > >
> > >
> > > 2018-05-13 22:29 GMT+02:00 Jakub Zawadzki <darkjames-ws () darkjames pl>:
> > >
> > > Hello,
> > >
> > > W dniu 2018-05-13 17:15, Eugène Adell napisał(a):
> > >
> > > I'm facing a problem on my development server (Ubuntu 16.04 hosted on
> > > VMWARE) when trying to compile Wireshark. It was working with older
> > > versions (2.0 for example), but now it's like the compilation will never
> > > end.
> > > I installed/updated all the required packages, since version 2.6 seems
> > > quite different.
> > >
> > > I'm doing an strace -f -o to find out what could be wrong, but no clue.
> > > The
> > > strace log being too big, here is how it looks like :
> > >
> > > 1900  execve("/usr/bin/cmake", ["cmake", "-LH", "../wireshark"], [/*
> > > 22 vars */]) = 0
> > > [cut]
> > >
> > >
> > >
> > > thousands of lines such as :
> > >
> > > 1901  open("/proc/537/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/538/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/539/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/540/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/541/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/542/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/543/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > > 1901  open("/proc/544/status", O_RDONLY) = -1 ENOENT (No such file or
> > > directory)
> > >
> > >
> > > then thousands and thousands of lines such as the following, and it seems
> > > it will never end :
> > >
> > > 1901  open("/proc/882/status", O_RDONLY) = 3
> > > 1901  fstat(3, {st_mode=S_IFREG|0444, st_size=0, ...}) = 0
> > > 1901  read(3, "Name:\tvmhgfs-fuse\nUmask:\t0000\nSt"..., 1024) = 1024
> > > 1901  read(3, "0000,00000000,00000000,00000000,"..., 1024) = 263
> > > 1901  read(3, "", 1024)                 = 0
> > > 1901  close(3)                          = 0
> > > 1901  open("/proc/965/status", O_RDONLY) = 3
> > > 1901  fstat(3, {st_mode=S_IFREG|0444, st_size=0, ...}) = 0
> > > 1901  read(3, "Name:\tsystemd-logind\nUmask:\t0022"..., 1024) = 1024
> > > 1901  read(3, "0000000,00000000,00000000,000000"..., 1024) = 269
> > > 1901  read(3, "", 1024)                 = 0
> > > 1901  close(3)                          = 0
> > > 1901  open("/proc/968/status", O_RDONLY) = 3
> > > 1901  fstat(3, {st_mode=S_IFREG|0444, st_size=0, ...}) = 0
> > > 1901  read(3, "Name:\tdbus-daemon\nUmask:\t0022\nSt"..., 1024) = 1024
> > > 1901  read(3, "00,00000000,00000000,00000000,00"..., 1024) = 283
> > > 1901  read(3, "", 1024)                 = 0
> > > 1901  close(3)                          = 0
> > >
> > >
> > > How can I resolve this ?
> > >
> > >
> > > Not sure, but it seems that first subprocess (1901 == 1900 + 1) makes
> > > some strange things.
> > >
> > > Looking on my strace output of cmake -LH ../wireshark:
> > >
> > > 6410  execve("/usr/bin/cmake", ["cmake", "-LH", "../wireshark/"],
> > > 0x7ffe72092520 /* 32 vars */) = 0
> > > (..)
> > > 6410  clone(child_stack=NULL, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD,
> > > child_tidptr=0x7f7cde41ba50) = 6411
> > > 6411  execve("/usr/bin/python", ["/usr/bin/python", "-c", "import sys;
> > > sys.stdout.write(';'.join([str(x) for x in sys.version_info[:3]]))"],
> > > 0x7ffc7adb4958 /* 32 vars */) = 0
> > >
> > > first execve() is /usr/bin/python -c "import sys;
> > > sys.stdout.write(';'.join([str(x) for x in sys.version_info[:3]]))"
> > >
> > > Is it same for you? If you run from same shell:
> > >   /usr/bin/python -c "import sys; sys.stdout.write(';'.join([str(x) for
> > > x in sys.version_info[:3]]))"
> > >
> > > does it hang?
> > >
> > > Could you please attach gzip compressed strace log?
> > >
> > >
> > > Kind Regards,
> > > Jakub.
> > >
> > >
> > > ____________________________________________________________
> > > _______________
> > > Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
> > > Archives:    https://www.wireshark.org/lists/wireshark-dev
> > > <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.wireshark.org_lists_wireshark-2Ddev&d=DwMFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=sylJsu9F3L2Fxzxkcz-j31k336lYCd03sKci0IgnttE&m=VyJzY90wLHzs7Hv2Xaqcj8u-pTGT9mDa_0FaNH-aiyA&s=qmxgFz9Mc1F23DLGlwWFlak3-5qavJqq-GK5QNoSyL0&e=>
> > > Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
> > > <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.wireshark.org_mailman_options_wireshark-2Ddev&d=DwMFaQ&c=uilaK90D4TOVoH58JNXRgQ&r=sylJsu9F3L2Fxzxkcz-j31k336lYCd03sKci0IgnttE&m=VyJzY90wLHzs7Hv2Xaqcj8u-pTGT9mDa_0FaNH-aiyA&s=N6_zXwmznWDlkEVq-UcnNPbsPjT8nqkJ2astIZS1efc&e=>
> > >             mailto:wireshark-dev-request () wireshark org?subject=
> > > unsubscribe
> > >
> > >
> > >
> > > ____________________________________________________________
> > > _______________
> > > Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
> > > Archives:    https://www.wireshark.org/lists/wireshark-dev
> > > Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
> > >              mailto:wireshark-dev-request () wireshark org?subject=
> > > unsubscribe
> >
> > ____________________________________________________________
> > _______________
> > Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
> > Archives:    https://www.wireshark.org/lists/wireshark-dev
> > Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
> >              mailto:wireshark-dev-request () wireshark org?subject=
> > unsubscribe
>
>
> ____________________________________________________________
> _______________
> Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
> Archives:    https://www.wireshark.org/lists/wireshark-dev
> Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
>              mailto:wireshark-dev-request () wireshark org?subject=
> unsubscribe
___________________________________________________________________________
Sent via:    Wireshark-dev mailing list <wireshark-dev () wireshark org>
Archives:    https://www.wireshark.org/lists/wireshark-dev
Unsubscribe: https://www.wireshark.org/mailman/options/wireshark-dev
             mailto:wireshark-dev-request () wireshark org?subject=unsubscribe