‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ On Wednesday, May 20, 2020 6:57 PM, Julian Andres Klode <j...@debian.org> wrote:
> Control: fixed -1 1.5~alpha4 > Control: tag -1 unreproducible > > On Wed, May 20, 2020 at 03:41:57PM +0000, meht...@protonmail.com wrote: > > > ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ > > On Wednesday, May 20, 2020 3:04 PM, Julian Andres Klode j...@debian.org > > wrote: > > > > > Control: tag -1 moreinfo > > > On Wed, May 20, 2020 at 02:41:12PM +0200, Mehturt wrote: > > > > > > > Package: apt-transport-https > > > > Version: 1.4.10 > > > > Severity: important > > > > Dear Maintainer, > > > > running "apt update" finishes with segmentation fault > > > > when the following entry is in my sources.list: > > > > deb http://packages.matrix.org/debian stretch main > > > > Removing the entry from sources.list produces no crash. > > > > This seems to be similar to already reported > > > > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=778375 > > > > > > Please obtain a full backtrace. > > > > Sorry, not sure what full backtrace means, but I installed > > apt-transport-https-dbgsym, > > run apt update again and got this using gdb and the core file: > > well you clearly miss a libcurl-gnutls dbgsym package, and you need to pass > "full" to bt. Any idea how to install this package? # apt install libcurl3-gnutls-dbgsym Reading package lists... Done Building dependency tree Reading state information... Done E: Unable to locate package libcurl3-gnutls-dbgsym # apt-cache search libcurl3-gnutls dbg > > > Core was generated by `/usr/lib/apt/methods/https.bin'. > > Program terminated with signal SIGSEGV, Segmentation fault. > > #0 0x00007f0d6c04f4d7 in ?? () from > > /usr/lib/x86_64-linux-gnu/libcurl-gnutls.so.4 > > (gdb) bt > > #0 0x00007f0d6c04f4d7 in ?? () from > > /usr/lib/x86_64-linux-gnu/libcurl-gnutls.so.4 > > #1 0x00005638ad9bf74b in HttpsMethod::~HttpsMethod (this=0x7ffe4abe9690, > > __in_chrg=<optimized out>) at ./methods/https.cc:538 > > #2 main (argv=<optimized out>) at ./methods/https.cc:546 > > This fails in curl_easy_cleanup(curl) call. > > It might need running this in valgrind to figure out what's going > on, as that seems like something that should work, and it might be > a corruption elsewhere. That said, no idea really how to run the > methods in valgrind. Sorry, so this bug is now tagged unreproducible, does it mean you can't reproduce it with apt-transport-https from stretch?
