Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Steven Levine

Pages: 1 2 3 [4] 5
46
Networking / Re: Samba groups are broken
« on: July 26, 2024, 01:09:43 am »
If you want to debug this on your system, here are a couple of ideas in case you have not already tried them.

If you grab a copy of the smbusers sources from

  http://svn.netlabs.org/repos/samba/trunk/guitools/smbusers

you might find it easier to follow the code path

The message

  Samba groups are broken, only 0 instead of 9 groups found

is defined at

  guitools\smbusers\rc\lang\smbusers_en.mkm:129
  SMU0112I: Samba groups are broken, only %1 instead of %2 groups found.

and generated for display at

guitools\smbusers\SMBUSERS.VRX:1278
        Msg.Text = NLVGetMessage(112, gline.0, gname.0)

The 0 in gline.0 is the result of the output of the net rpc command invoked at:

guitools\smbusers\SMBUSERS.VRX:1231
    address cmd samba.!netexe' rpc group -l 'DebugLevel' 'UserCred' 1>'grouplistf' 2>'samba.!error

The code that follows parses the content of the "grouplistf" file and sets gline.0

The grouplistf file name is defined at

guitools\smbusers\SMBUSERS.VRX:1227
    grouplistf = TempDir'netgrouplist.txt'

The content of this file might give you some ideas as to what is not working correctly.  You are either getting no output from the net rpc command or output that the code does not understand.

If the file content is not helpful, mapping

    address cmd samba.!netexe' rpc group -l 'DebugLevel' 'UserCred' 1>'grouplistf' 2>'samba.!error

to something you can run from the command line might give you some useful information.  You can start with

  net rpc group -l

and add options as you go.

The samba.!error file might also contain some useful info.  The file is generated by:

guitools\shared\cltinit.vrs:205
    samba.!error = SysTempFileName(TempDir'smb_err.???')

so look for the newest smb_err.* file in your temp directory.


47
Programming / Re: GCC - updates
« on: July 25, 2024, 10:44:55 pm »
If I had to guess, I suspect the changed code is a side effect of the removal of .stabs support.  Without the .stabs entry, there is no obvious way to reference what appears to be a static constructor.

In the 13x, it appears that the static constructor is gone, probably because the code generator has be updated or possibly because the default compiler options have changed.

The exceptq report does not contain sufficient dis-assembly for me to see how the address in EBX is calculated.  Please upload a version of xcpt.exe with debug debug and I'll see if I can figure out where the bad address is coming from.

48
Networking / Re: Updated Samba Client
« on: July 13, 2024, 09:51:05 pm »
David,

Most of the time, Ctrl-Alt-F10-F10 will bring up the system dump prompts, even if the keyboard is not responding to normal keystrokes.  Whether or not you will be able to capture a system dump, depends on your boot environment, much more than it used to.

Depending on what starting smbd breaks, you still might be able to capture a process dump.

 - open a command line session where you will start smbd
 - open a 4os2 session where you will manage the process dump facility
 - configure the process dump facility with something like
      pdumpctl n f v smbd
     delay 30 & pdumpctl f smbd
 - switch to the smbd command line session and start smbd

If the system is sufficiently alive after smbd starts, pdumpctl will force a process dump after 30 seconds


49
Networking / Re: Updated Samba Client
« on: July 13, 2024, 02:39:00 am »
David,

Can I assume that CAD failed to pop up CADH?

If you can get CADH to pop up, you can usually capture a process dump, even if you cannot kill the app that is hung.

If CADH will not pop up, you could try to capture a system dump.  This would likely tell us where smbd is going astray.

Steven

50
Applications / Re: General Software Testing
« on: June 30, 2024, 05:51:59 am »
Fractz could be fixed to work with systems with more than 256 colors.  All that is needed is to avoid array overflow.  It might be sufficient to increase

static LONG blTable[256],   // color tables
            alTable[256];

to hold 16M entries.  If this makes the program too big to fit in memory, there are other options.


51
Applications / Re: General Software Testing
« on: June 25, 2024, 11:19:57 pm »

The executable included in the zip file will never run on a typical modern system.  It's limited to 256 colors.  With more it will eventually trap at or near

FRACTZ.CPP:1369
             pbmi->argbColor.bBlue  =  blTable & 0x000000ff;

because blTable is defined as

FRACTZ.CPP:246
static LONG blTable[256],   // color tables
            alTable[256];

You should be able to get FRACTZ.EXE to build with VAC 3.0.8 using nmake and tweaking the included MAKEFILE.

Since I wanted to see where the trap occurred, I built for debug with:

MAKEFILE:17
FRACTZ.EXE:  \
  fractz.OBJ \
  FRACTZ.RES \
  MAKEFILE
   ICC.EXE @<<
 /B " /pmtype:pm /de"
 /Fe"FRACTZ.EXE"
 fractz.OBJ fractz.def
<<KEEP

and

MAKEFILE:35
{.}.cpp.obj:
   ICC.EXE /J- /Gm /Gf /Ti /C   .\$*.cpp

My testing says stack size had nothing to do with the exception.  However, it
you find you really need a larger stack to support more colors, you can set
the stack size in the .def file with:

FRACTZ.DEF:9
STACKSIZE 0x30000

I'm not sure why your link attempt went wrong.  Probably the linker you attempted to use.  I can link here with either icc.exe or 3.0.8's ilink.


52
Applications / Re: General Software Testing
« on: May 22, 2024, 06:10:29 pm »
This appears to be an older version of the IBM supplied migration tool (\os2\install\migrate.exe).

53
General Discussion / Re: smp.inf - Adendum
« on: May 09, 2024, 03:42:05 am »
To keep this interesting, there is a 3rd version of the November, 1996 smp.inf.  This one adds the definition for the CVK_CMD_SCANPCB packet command.

54
Programming / Re: ClassiCube porting thread
« on: May 06, 2024, 02:30:18 am »
Exception codes are documented in the OS/2 Debugging Handbook (sg244640.inf) and the Control Program Guide and Reference (GPG&R)

Both annotate the value differently.  Given c0000005, you need to search for 0c0000005h in the Debugging Handbook and for 0xc0000005 in the CPG&R.

As Dave noted, you need to expand the per thread stack for TID 0xa.  You can use LIBC_THREAD_MIN_STACK_SIZE to set the minimum at runtime or, assuming that ClassiCube is using pthreads, you can adjust the compile time values by modifying the code that eventually invokes pthread_attr_setstacksize()

Note that while you can get stack overflow if the stack is too small, you can also get stack overflow if the code recurses unexpectedly.  A popuplog is typically not going to tell why the stack overflow happened.  This requires an exceptq report or a process dump file.

BTW, are your sources available anywhere?

55
Programming / Re: ClassiCube porting thread
« on: May 05, 2024, 03:17:28 am »
This looks like stack overflow in a secondary thread (TID 0xa).  If possible, try rebuilding with larger thread stacks.

Steven

56
Hardware / Re: OpenWatcom Discussion
« on: February 02, 2024, 02:37:21 am »
When talking about repositories, current is a subjective term.  The same is true for which is best.

The OpenWatcom repository Gregg and I maintain was intended as a placeholder until Michal got openwatcom.org back up an running.  The lack of an issues tracker was an oversight, which has been corrected. Silvan is the first to mention it.

Gregg and I had commits that needed to be made, so the repo came to be.

The repository is a fork of Jiri Malak's https://github.com/open-watcom/owp4v1copy, which is about 3 years old.
Jiri's owp4v1copy was current as of when the Perforce repository went away.

Jiri's owp4v1copy is significantly newer than iainnicol's.

Jiri's OpenWatcom repositories have been around for a while.  He created them back in 2012.

57
Networking / Re: Samba Users and Groups
« on: January 16, 2024, 05:45:16 pm »
Sorry if I ruffled any feathers.

For those that might be curious, assuming you are running an updated system, there is

  %UNIXROOT%\usr\share\doc\libcx-0.7.3\CHANGELOG.md
  %UNIXROOT%\usr\share\doc\libcx-0.7.3\README.md

and for libc, there is

  %UNIXROOT%\usr\share\doc\libc-0.1.12\CHANGELOG.md
  %UNIXROOT%\usr\share\doc\libc-0.1.12\README.md

I will return to my cave.

58
Networking / Re: Samba Users and Groups
« on: January 15, 2024, 09:53:39 am »
I generally ignore exceptq reports generated without proper symbolics.  The exceptq documentation explains clearly what is required.

Also, by now it should be well known that when libcx0 traps it's common sense to look in %UNIXROOT/var/log/libcx in
case a libcx log was generated.

However, this libcx trap location is well known and the exceptq report happened to include a somewhat readable version of part of the log:

 0062F160 : 1E061F79 00000085 1E0625FC 1E061F8C : y........%......
 0062F170 : 0062F1CD 1E061F8C 00000000 0062F1EC : ..b...........b.
 0062F180 : 30303030 33616664 20313020 74727341 : 0000dfa3 01 Asrt
 0062F190 : 72777020 2E657469 33313A63 72703A33 :  pwrite.c:133:pr
 0062F1A0 : 5F646165 69727770 203A6574 65737341 : ead_pwrite: Asse
 0062F1B0 : 6F697472 6166206E 64656C69 7261203A : rtion failed: ar
 0062F1C0 : 3D3D2063 5F4F4E20 4F525245 36282052 : c == NO_ERROR (6
 0062F1D0 : 3D0A0A29 3D3D3D3D 42494C20 72207843 : )..===== LIBCx r

For some, as yet unknown reason, the mutex that serializes pread/pwrite operations has gone bad.

This smbpasswd.exe is a tad antique.  I recommend replacing it with one of Paul's new builds and see if this is sufficient to make the exception go away.

59
Applications / Re: ArcaOS 5.1 - First Impressions
« on: January 04, 2024, 02:45:15 am »
I cannot replicate Lars' failure mode either.

If you want to use top to monitor threads, there is an easier way.

Use the Display menu to select Process Details so the thread count is always visible.  Use the Sort menu to sort by Name so that the process list does not change so much.  You might want to increase the refresh rate with Options ->
Refresh Rate.

Lars, when you say many, how many is many?  I tested with a folder containing 3000+ files.

60
Programming / Re: Git push to Github issue.
« on: December 29, 2023, 09:17:05 am »
403 means you probably still have an authentication issue.  I would suspect a push URL issue.  The rules have changed over the years.  For example, the git: scheme is no longer allowed.  The same is true for plain http:.

Check the push URL defined in the repo's config file (i.e.  .git\config)

These days, I usually use something like:

  url = https://StevenLevine@github.com/StevenLevine/php-os2.git

There's rarely a need for a specific pushurl.

Can you push to any repo?

A git push --dry-run to your repo fails here as expected.  I have a known good token, but no permission to push to the repo.

FWIW, I maintain a sandbox repo for working out these kinds of issues.  I can tune the access permissions without messing with the production repos.

Pages: 1 2 3 [4] 5