Great, thank you very much. The fix works for me :)

Sorry I do not understand what you try to say with the sentence: "The bug you reported is not in this release (never was there)".
If I download the released version from your download page, then the source code is exactly what I reported in the original bug report. Even the compiler complains with a warning about that line:

umount_davfs.c: In function ‘main’:
umount_davfs.c:153:33: warning: format ‘%a’ expects argument of type ‘float *’, but argument 3 has type ‘char **’ [-Wformat=]
  153 |     if (!file || fscanf(file, "%a[0-9]", &pid) != 1 || !pid) {
      |                                ~^        ~~~~
      |                                 |        |
      |                                 float *  char **

I will also attach a strace output for you.


(file #47831)

comment #1:

> I also get segmentation fault when umounting my egroupware webdav share since updating to davfs2 1.5.5.
>
> I use the following command to mount:
> sudo mount -t davfs -o uid=myname,gid=users https://path/to/webdav/ /home/myname/Documents/mountpoint/
>
>

Just opened an account, so you can request any additional info if needed. I am running the client on KUbuntu 19.10, davfs2 version is the one packaged from the Ubuntu repositories, the server is egroupware.

Reading of the PID is wrong. This snippet from umount_davfs.c is not doing what it is supposed to do:
151     char *pid = NULL;
152     FILE *file = fopen(pidfile, "r");
153     if (!file || fscanf(file, "%a[0-9]", &pid) != 1 || !pid) {

You are saving a float bit representation into a char** (it fits, but then the char* is pointing to random memory, i.e. the PID)

In CVS I can see that this was fixed in CVS revision 1.7, but I do not really get, how the relation of the CVS repo is with the released versions, since the released versions seem to be coming from branches and never get the fixes from HEAD?