skibbidypaps

[skibbidypaps]

I’m happy to run more tests with a new package. Is there some way my account can be set to no longer require moderation? I’m clearly acting in good faith (not a spammer) and it adds a long delay in our conversation.

 

Please let me know when it has been sent.

[skibbidypaps]

Yes, the problem persists in the debugging package. The debugging package just appears to just have all the debugging options turned on (refer to https://knowledgebase.nomachine.com/DT11R00182)

It wasn’t an actual new release – so I wasn’t expecting a resolution to my issue. And when I tested it, the issue still presented itself in the package I was sent.

[skibbidypaps]

The tag currently associated with this is a bit of a misnomer, and since I had more to add it was the perfect opportunity to add a tag.

Under Mac OS, the uid_t type is an unsigned 32-bit int: /Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/usr/include/sys/_types/_uid_t.h:

#ifndef _UID_T

#define _UID_T

#include <sys/_types.h> /* __darwin_uid_t */

typedef __darwin_uid_t        uid_t;

#endif  /* _UID_T */

 

And from /usr/include/sys/_types.h:

typedef __uint32_t      __darwin_id_t;          /* [XSI] pid_t, uid_t, or gid_t*/

I wrote a little test program that uses getuid() to get the pid of the running process and then uses both getpwnam (for my user) and getpwuid(). It prints out the entries from the returned passwd structure just fine for both calls. I repeated this for root and it too works just fine.

It appears as though all but one of the running NoMachine processes are running as my local user, and the outlier is nxexec.

I’m not sure where NoMachine’s code has gone wrong, and there are many interfaces to these syscalls that are language dependent, but I’m able to retrieve the data using the most obvious and POSIX compliant interfaces.

#include <sys/types.h>

#include <uuid/uuid.h>

#include <pwd.h>

#include <stdio.h>

#include <unistd.h>

 

int

main(int argc, char *argv[]) {

 

uid_t this_uid = 0;

struct passwd *u_mypw, *i_mypw;

char *who = “root”;

uid_t whos_uid = 0;

 

this_uid = getuid();

u_mypw = getpwnam(who);

i_mypw = getpwuid(whos_uid);

 

printf(“uid is %d\n”, this_uid);

printf(“%s’s UID is %d via getpwnam()\n”, who, u_mypw->pw_uid);

printf(“UID %d’s name is %s via getpwuid()\n”, whos_uid, i_mypw->pw_name);

}

$ ./test

uid is 1974106972

root’s UID is 0 via getpwnam()

UID 0’s name is root via getpwuid()

 

[skibbidypaps]

Just send the logs on to you via email, Britgirl.

[skibbidypaps]

Received and installed, will follow-up with logs.

[skibbidypaps]

Hi,

A gmail search for in:all nomachine debug doesn’t return any results in my inbox. A search for in:all nomachine only returns results related to this thread.

Would you please share the subject and/or from address of the email that sent the debugging packages?

 

 

[skibbidypaps]

Happy to run a debug version and assist in debugging it. One problem you may have in reproducing it is having Macs configured to resemble corporate deployments with various profiles.

In other posts related to this issue (which I believe were prematurely closed) the uid was similarly a large integer that indicated active directory users

[skibbidypaps]

This might only be an error for me, but I’m unable to load the error in my browser on the forums. Attaching it once again here. And hoping for some replies to my thorough post.

Attachments:

1. 

   nomachine_error.jpg
   

[skibbidypaps]

To clarify what’s written above, when using the terms client and server that refers to the use of NoMachine. The Mac is not running a Mac OS server operating system.

[Author]

Posts