• ctfconvert errors during buildworld

    From Garrett Wollman@wollman@bimajority.org to muc.lists.freebsd.stable on Wed Oct 21 23:45:16 2026
    From Newsgroup: muc.lists.freebsd.stable

    I have noticed for quite some time a stream of errors from `make -s buildworld`, all of the form:

    ERROR: ctfconvert: failed to initialize DWARF: (null) [_dwarf_elf_decompress(235)]

    The errors don't stop the build so I guess they're not really errors,
    but what actually is the cause? If I build without `-s` I can see
    that they are as expected emitted by the ctfconvert(1) utility which
    runs on each object file after it is emitted by the compiler.

    Occasionally there will be a

    WARNING: ctfmerge: No ctf sections found to merge

    which I guess is the result of ctfconvert having failed to do anything
    useful during the compilation phase, but why?

    -GAWollman



    --
    Posted automagically by a mail2news gateway at muc.de e.V.
    Please direct questions, flames, donations, etc. to news-admin@muc.de
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Adrian Chadd@adrian@freebsd.org to muc.lists.freebsd.stable on Mon Oct 26 16:03:43 2026
    From Newsgroup: muc.lists.freebsd.stable

    --00000000000027df04065ec65830
    Content-Type: text/plain; charset="UTF-8"

    I've noticed it to on -head. I think we should file a bug and chase it up.


    -a


    On Wed, 21 Oct 2026 at 20:45, Garrett Wollman <wollman@bimajority.org>
    wrote:

    I have noticed for quite some time a stream of errors from `make -s buildworld`, all of the form:

    ERROR: ctfconvert: failed to initialize DWARF: (null) [_dwarf_elf_decompress(235)]

    The errors don't stop the build so I guess they're not really errors,
    but what actually is the cause? If I build without `-s` I can see
    that they are as expected emitted by the ctfconvert(1) utility which
    runs on each object file after it is emitted by the compiler.

    Occasionally there will be a

    WARNING: ctfmerge: No ctf sections found to merge

    which I guess is the result of ctfconvert having failed to do anything
    useful during the compilation phase, but why?

    -GAWollman




    --00000000000027df04065ec65830
    Content-Type: text/html; charset="UTF-8"
    Content-Transfer-Encoding: quoted-printable

    <div dir=3D"ltr"><div>I&#39;ve noticed it to on -head. I think we should fi=
    le a bug and chase it up.</div><div><br></div><div><br></div><div>-a</div><= div><br></div></div><br><div class=3D"gmail_quote gmail_quote_container"><d=
    iv dir=3D"ltr" class=3D"gmail_attr">On Wed, 21 Oct 2026 at 20:45, Garrett W= ollman &lt;<a href=3D"mailto:wollman@bimajority.org">wollman@bimajority.org= </a>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:= 0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">=
    I have noticed for quite some time a stream of errors from `make -s<br> buildworld`, all of the form:<br>

    ERROR: ctfconvert: failed to initialize DWARF: (null) [_dwarf_elf_decompres= s(235)]<br>

    The errors don&#39;t stop the build so I guess they&#39;re not really error= s,<br>
    but what actually is the cause?=C2=A0 If I build without `-s` I can see<br> that they are as expected emitted by the ctfconvert(1) utility which<br>
    runs on each object file after it is emitted by the compiler.<br>

    Occasionally there will be a <br>

    WARNING: ctfmerge: No ctf sections found to merge<br>

    which I guess is the result of ctfconvert having failed to do anything<br> useful during the compilation phase, but why?<br>

    -GAWollman<br>


    </blockquote></div>

    --00000000000027df04065ec65830--


    --
    Posted automagically by a mail2news gateway at muc.de e.V.
    Please direct questions, flames, donations, etc. to news-admin@muc.de
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Garrett Wollman@wollman@bimajority.org to muc.lists.freebsd.stable on Fri Oct 30 16:41:45 2026
    From Newsgroup: muc.lists.freebsd.stable

    <<On Mon, 26 Oct 2026 16:03:43 -0700, Adrian Chadd <adrian@freebsd.org> said:

    I've noticed it to on -head. I think we should file a bug and chase it up.

    I'd be happy to sign on to that. I don't usually build without `-j N`
    or `-s` but I can certainly reproduce the error regardless. But it
    would be more helpful if someone who regularly builds from main can do
    so.

    -GAWollman



    --
    Posted automagically by a mail2news gateway at muc.de e.V.
    Please direct questions, flames, donations, etc. to news-admin@muc.de
    --- Synchronet 3.22a-Linux NewsLink 1.2