[Patch] - Fix for bug #2558, InitDB failed to run on windows 2003

[dror bar]

Some parts of this message have been removed. Learn more about Nabble's security policy.

Hi All,

 

 

As some of you already noticed and as mentioned in pgsql-hackers  before,

the initDB process failed to run on some windows 2003 machines.

 

Although this issue was already discussed by James Hughes,Martijn,Jim Nasby and others I didn't no patch was created and

no fix was committed to the CVS.

 

Description:

 

On some windows 2003 machine there is no access to the nul device for non administrator's users.

The initDB.c redirect the output to > DEVNUL which on windows defined has "nul".

 

There were two options to solve this issue:

1. Create a new file , grant a write permission for the Postgres user  and redirect the output to that file. (EnterpriseDB  use this method)
2. Canceling the redirection at all.

 

I choose the second option and omit the redirection in any case that it windows machine and the redirection was sent to DEVNULL.

 

The only files that I changed are: initDB.c, exec.c and pg_ctl.c

 

 

I think postgres must solve this issue, as more and more windows 2003 machine will be discovered with this limitation which prevent from the users to install (actually to initialize ) the DB.

 

One more comment:

As today we have  and MSI installer which redirect the initDB output to a log file by default (and we can also run it in the background) there is no point to use redirection at all.

 

The fixed files are attached.

 

Regards

Dror Bar-Gil

 

 

 

---

With MSN Spaces email straight to your blog. Upload jokes, photos and more. It's free! It's free!


---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

               http://archives.postgresql.org

[Alvaro Herrera-7]

dror wrote:

> There were two options to solve this issue:
>
> Create a new file , grant a write permission for the Postgres user
> and redirect the output to that file. (EnterpriseDB  use this method)
> Canceling the redirection at all.
>  
> I choose the second option and omit the redirection in any case that
> it windows machine and the redirection was sent to DEVNULL.
>  
> The only files that I changed are: initDB.c, exec.c and pg_ctl.c

Please submit the changes as patches, instead of the whole files.  Also,
please specify which branch do these patches apply -- is this for 8.1,
or for the current development code?  When checked against the 8.1
pg_ctl.c, the file you sent only contains a regression for a bug fix,
and no change related to what you describe above.

On the other hand, it may be useful to lose the redirection only on the
cases where it fails, so we still have reasonable behavior on non-broken
platforms.  Or maybe there's a better solution.

--
Alvaro Herrera                                http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support

---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

[Andrew Dunstan]

Alvaro Herrera wrote:

> dror wrote:
>
>  
>> There were two options to solve this issue:
>>
>> Create a new file , grant a write permission for the Postgres user
>> and redirect the output to that file. (EnterpriseDB  use this method)
>> Canceling the redirection at all.
>>  
>> I choose the second option and omit the redirection in any case that
>> it windows machine and the redirection was sent to DEVNULL.
>>  
>> The only files that I changed are: initDB.c, exec.c and pg_ctl.c
>>    
>
> Please submit the changes as patches, instead of the whole files.  Also,
> please specify which branch do these patches apply -- is this for 8.1,
> or for the current development code?  When checked against the 8.1
> pg_ctl.c, the file you sent only contains a regression for a bug fix,
> and no change related to what you describe above.
>
> On the other hand, it may be useful to lose the redirection only on the
> cases where it fails, so we still have reasonable behavior on non-broken
> platforms.  Or maybe there's a better solution.
>
>  

I am inclined to say we should make it into a runtime test and use a
tmpfile on Windows if the test fails. I am more than somewhat perplexed
as to why the NUL device should be a security risk ... what are they
thinking??

The case that bothers me more is where input is redirected - will that
also work?

cheers

andrew

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
       subscribe-nomail command to majordomo@... so that your
       message can get through to the mailing list cleanly

[Tom Lane-2]

Andrew Dunstan <andrew@...> writes:
> I am more than somewhat perplexed as to why the NUL device should be a
> security risk ... what are they thinking??

Frankly, I don't believe it; even Microsoft can't be that stupid.
And I can't find any suggestion that they've done this in a google
search.  I think the OP is misdiagnosing his problem.

                        regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
       choose an index scan if your joining column's datatypes do not
       match

[dror bar]

Some parts of this message have been removed. Learn more about Nabble's security policy.

1. When saying:
   "Please submit the changes as patches, instead of the whole files".  
   Do you mean to send each file seperately? or other issues as well?
2. The change was test and design for 8.1.14, but as far as I see it is also true for any other version.
   Of course merge is needed in case that the files were changed since then, however , due to the fact that it is only few rows it will be easy to do it.
3. Alvaro wrote:
   "it may be useful to lose the redirection only on the  cases where it fails, so we still have reasonable behavior on non-broken
    platforms"
   
   Nice idea, but it is really doesn't matter:  on other platform than win32, the code left unchanged! the fix is only relevant for win32 on which (as I explained) the MSI installer (if used) redirect the output by default to a log file.

Regards
Dror

---

> Date: Mon, 14 Aug 2006 12:40:25 -0400

> From: alvherre@...
> To: dror_b@...
> CC: pgsql-patches@...
> Subject: Re: [PATCHES] [Patch] - Fix for bug #2558, InitDB failed to run on windows 2003
>
> dror wrote:
>
> > There were two options to solve this issue:
> > 
> > Create a new file , grant a write permission for the Postgres user
> > and redirect the output to that file. (EnterpriseDB  use this method)
> > Canceling the redirection at all.
> >  
> > I choose the second option and omit the redirection in any case that
> > it windows machine and the redirection was sent to DEVNULL.
> >  
> > The only files that I changed are: initDB.c, exec.c and pg_ctl.c
>
> Please submit the changes as patches, instead of the whole files.  Also,
> please specify which branch do these patches apply -- is this for 8.1,
> or for the current development code?  When checked against the 8.1
> pg_ctl.c, the file you sent only contains a regression for a bug fix,
> and no change related to what you describe above.
>
> On the other hand, it may be useful to lose the redirection only on the
> cases where it fails, so we still have reasonable behavior on non-broken
> platforms.  Or maybe there's a better solution.
>
> -- 
> Alvaro Herrera                                http://www.CommandPrompt.com/
> PostgreSQL Replication, Consulting, Custom Development, 24x7 support

---

Express yourself instantly with Windows Live Messenger! Windows Live Messenger!

[Andrew Dunstan]

1. a patch is generated by the program "diff"
2. before we do anything, as Tom Lane says, we need verification of the
problem, preferably in writing from Microsoft.

cheers

andrew


dror wrote:

>
>   1.
>       When saying:
>       "Please submit the changes as patches, instead of the whole files".  
>       Do you mean to send each file seperately? or other issues as well?
>   2.
>       The change was test and design for 8.1.14, but as far as I see
>       it is also true for any other version.
>       Of course merge is needed in case that the files were changed
>       since then, however , due to the fact that it is only few rows
>       it will be easy to do it.
>   3.
>       Alvaro wrote:
>       "it may be useful to lose the redirection only on the
>        cases where it fails, so we still have reasonable behavior on non-broken
>        platforms"
>
>       Nice idea, but it is really doesn't matter:  on other platform
>       than win32, the code left unchanged! the fix is only relevant
>       for win32 on which (as I explained) the MSI installer (if used)
>       redirect the output by default to a log file.
>
> Regards
> Dror
> ------------------------------------------------------------------------
>
> > Date: Mon, 14 Aug 2006 12:40:25 -0400
> > From: alvherre@...
> > To: dror_b@...
> > CC: pgsql-patches@...
> > Subject: Re: [PATCHES] [Patch] - Fix for bug #2558, InitDB failed to
> run on windows 2003
> >
> > dror wrote:
> >
> > > There were two options to solve this issue:
> > >
> > > Create a new file , grant a write permission for the Postgres user
> > > and redirect the output to that file. (EnterpriseDB  use this method)
> > > Canceling the redirection at all.
> > >  
> > > I choose the second option and omit the redirection in any case that
> > > it windows machine and the redirection was sent to DEVNULL.
> > >  
> > > The only files that I changed are: initDB.c, exec.c and pg_ctl.c
> >
> > Please submit the changes as patches, instead of the whole files.  Also,
> > please specify which branch do these patches apply -- is this for 8.1,
> > or for the current development code?  When checked against the 8.1
> > pg_ctl.c, the file you sent only contains a regression for a bug fix,
> > and no change related to what you describe above.
> >
> > On the other hand, it may be useful to lose the redirection only on the
> > cases where it fails, so we still have reasonable behavior on non-broken
> > platforms.  Or maybe there's a better solution.
> >
> > --
> >
> Alvaro Herrera                                http://www.CommandPrompt.com/
> > PostgreSQL Replication, Consulting, Custom Development, 24x7 support
>
>
> ------------------------------------------------------------------------
> Express yourself instantly with Windows Live Messenger! Windows Live
> Messenger!
> <http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=joinmsncom/messenger>

---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

               http://archives.postgresql.org

[dror bar]

Some parts of this message have been removed. Learn more about Nabble's security policy.

Hi Andrew,
 
Regarding to your comments:
 
> 1. a patch is generated by the program "diff"
I will do it ,if needed

> 2. before we do anything, as Tom Lane says, we need verification of the 
> problem, preferably in writing from Microsoft.
I do understand that, but,  de-facto, the current implementation does not work, canceling the redirection (or open a log file) is not a matter of changing the OS behavior, therefore I don't see  why a formal verification from Microsoft is needed.
When this issue will be revealed in more and more system, it can be harmless to postgress reputation and critical problems for the end users.
In addition to the above,  as James Hughes have already mention before at item #2268 (http://archives.postgresql.org/pgsql-hackers/2006-03/msg00012.php):
He tried to get answers from microsoft but didn't get any respond.
 
Anyway,
 
For my own version, I solve the issue and built a private version, if you still want me to publish the patch (Just in case) I will do the diff and the effort needed, I don't think that Microsoft respond is needed in this case because it is implementation decision.
 
Regards;
 
Dror

---

> Date: Mon, 14 Aug 2006 17:42:58 -0400

> From: andrew@...
> To: dror_b@...
> CC: alvherre@...; pgsql-patches@...
> Subject: Re: [PATCHES] [Patch] - Fix for bug #2558, InitDB failed to run
>
>
>
> 1. a patch is generated by the program "diff"
> 2. before we do anything, as Tom Lane says, we need verification of the 
> problem, preferably in writing from Microsoft.
>
> cheers
>
> andrew
>
>
> dror wrote:
> >
> >   1.
> >       When saying:
> >       "Please submit the changes as patches, instead of the whole files".  
> >       Do you mean to send each file seperately? or other issues as well?
> >   2.
> >       The change was test and design for 8.1.14, but as far as I see
> >       it is also true for any other version.
> >       Of course merge is needed in case that the files were changed
> >       since then, however , due to the fact that it is only few rows
> >       it will be easy to do it.
> >   3.
> >       Alvaro wrote:
> >       "it may be useful to lose the redirection only on the
> >        cases where it fails, so we still have reasonable behavior on non-broken
> >        platforms"
> >
> >       Nice idea, but it is really doesn't matter:  on other platform
> >       than win32, the code left unchanged! the fix is only relevant
> >       for win32 on which (as I explained) the MSI installer (if used)
> >       redirect the output by default to a log file.
> >
> > Regards
> > Dror
> > ------------------------------------------------------------------------
> >
> > > Date: Mon, 14 Aug 2006 12:40:25 -0400
> > > From: alvherre@...
> > > To: dror_b@...
> > > CC: pgsql-patches@...
> > > Subject: Re: [PATCHES] [Patch] - Fix for bug #2558, InitDB failed to 
> > run on windows 2003
> > >
> > > dror wrote:
> > >
> > > > There were two options to solve this issue:
> > > > 
> > > > Create a new file , grant a write permission for the Postgres user
> > > > and redirect the output to that file. (EnterpriseDB  use this method)
> > > > Canceling the redirection at all.
> > > >  
> > > > I choose the second option and omit the redirection in any case that
> > > > it windows machine and the redirection was sent to DEVNULL.
> > > >  
> > > > The only files that I changed are: initDB.c, exec.c and pg_ctl.c
> > >
> > > Please submit the changes as patches, instead of the whole files.  Also,
> > > please specify which branch do these patches apply -- is this for 8.1,
> > > or for the current development code?  When checked against the 8.1
> > > pg_ctl.c, the file you sent only contains a regression for a bug fix,
> > > and no change related to what you describe above.
> > >
> > > On the other hand, it may be useful to lose the redirection only on the
> > > cases where it fails, so we still have reasonable behavior on non-broken
> > > platforms.  Or maybe there's a better solution.
> > >
> > > -- 
> > > 
> > Alvaro Herrera                                http://www.CommandPrompt.com/
> > > PostgreSQL Replication, Consulting, Custom Development, 24x7 support
> >
> >
> > ------------------------------------------------------------------------
> > Express yourself instantly with Windows Live Messenger! Windows Live 
> > Messenger! 
> > <http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=joinmsncom/messenger>
>

---

With MSN Spaces email straight to your blog. Upload jokes, photos and more. It's free! It's free!

Tom Lane wrote:

> Andrew Dunstan <andrew@...> writes:
>  
>> I am more than somewhat perplexed as to why the NUL device should be a
>> security risk ... what are they thinking??
>>    
>
> Frankly, I don't believe it; even Microsoft can't be that stupid.
> And I can't find any suggestion that they've done this in a google
> search.  I think the OP is misdiagnosing his problem.
>  

An older message suggests that a service pack induced this problem, per
MS. I just tried it as non-admin on a W2K3 machine with recent hotfixes,
and the command "dir >nul" _did_ work for me.
Though neglected, it still sounds like a virus scanner issue to me.

Regards,
Andreas


---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

               http://www.postgresql.org/docs/faq

[Bruce Momjian-5]

Andreas Pflug wrote:

> Tom Lane wrote:
> > Andrew Dunstan <andrew@...> writes:
> >  
> >> I am more than somewhat perplexed as to why the NUL device should be a
> >> security risk ... what are they thinking??
> >>    
> >
> > Frankly, I don't believe it; even Microsoft can't be that stupid.
> > And I can't find any suggestion that they've done this in a google
> > search.  I think the OP is misdiagnosing his problem.
> >  
> An older message suggests that a service pack induced this problem, per
> MS. I just tried it as non-admin on a W2K3 machine with recent hotfixes,
> and the command "dir >nul" _did_ work for me.
> Though neglected, it still sounds like a virus scanner issue to me.

Yes, it seems we will need more information on this.  We need someone at
a win32 command prompt to show us a "> nul" failure.

--
  Bruce Momjian   bruce@...
  EnterpriseDB    http://www.enterprisedb.com

  + If your life is a hard drive, Christ can be your backup. +

---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

               http://www.postgresql.org/docs/faq

Bruce Momjian wrote:

> Andreas Pflug wrote:
>> Tom Lane wrote:
>>> Andrew Dunstan <andrew@...> writes:
>>>  
>>>> I am more than somewhat perplexed as to why the NUL device should be a
>>>> security risk ... what are they thinking??
>>>>    
>>> Frankly, I don't believe it; even Microsoft can't be that stupid.
>>> And I can't find any suggestion that they've done this in a google
>>> search.  I think the OP is misdiagnosing his problem.
>>>  
>> An older message suggests that a service pack induced this problem, per
>> MS. I just tried it as non-admin on a W2K3 machine with recent hotfixes,
>> and the command "dir >nul" _did_ work for me.
>> Though neglected, it still sounds like a virus scanner issue to me.
>
> Yes, it seems we will need more information on this.  We need someone at
> a win32 command prompt to show us a "> nul" failure.

OTOH,
what issues might arise if the output is redirected to a legal tmp file?

Regards,
Andreas

---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
       choose an index scan if your joining column's datatypes do not
       match

[Bruce Momjian-5]

Andreas Pflug wrote:

> Bruce Momjian wrote:
> > Andreas Pflug wrote:
> >> Tom Lane wrote:
> >>> Andrew Dunstan <andrew@...> writes:
> >>>  
> >>>> I am more than somewhat perplexed as to why the NUL device should be a
> >>>> security risk ... what are they thinking??
> >>>>    
> >>> Frankly, I don't believe it; even Microsoft can't be that stupid.
> >>> And I can't find any suggestion that they've done this in a google
> >>> search.  I think the OP is misdiagnosing his problem.
> >>>  
> >> An older message suggests that a service pack induced this problem, per
> >> MS. I just tried it as non-admin on a W2K3 machine with recent hotfixes,
> >> and the command "dir >nul" _did_ work for me.
> >> Though neglected, it still sounds like a virus scanner issue to me.
> >
> > Yes, it seems we will need more information on this.  We need someone at
> > a win32 command prompt to show us a "> nul" failure.
>
> OTOH,
> what issues might arise if the output is redirected to a legal tmp file?

No idea, but we aren't going to change the code without more facts.  We
don't have the resources to be making code changes without concrete
information.

--
  Bruce Momjian   bruce@...
  EnterpriseDB    http://www.enterprisedb.com

  + If your life is a hard drive, Christ can be your backup. +

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
       subscribe-nomail command to majordomo@... so that your
       message can get through to the mailing list cleanly

[Tom Lane-2]

Andreas Pflug <pgadmin@...> writes:
> what issues might arise if the output is redirected to a legal tmp file?

Well, (1) finding a place to put the temp file, ie a writable directory;
(2) ensuring the file is removed afterwards; (3) not exposing the user
to security hazards due to unsafe use of a temp file (ye olde
overwrite-a-symlink risk).  Perhaps a few more I didn't think of.

It's not a trivial change, and the evidence presented so far hasn't
convinced me that we need to put in the effort.

                        regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend

[SenTnel]

Sorry, Im not an expert, and I have the same win 2003 server installation problem, but dont know what to do with the tree .c files downloaded as a patch, can you please direct me on how to use the patch?
Thanks !

Andrew Dunstan wrote:

1. a patch is generated by the program "diff"
2. before we do anything, as Tom Lane says, we need verification of the
problem, preferably in writing from Microsoft.

cheers

andrew


dror wrote:
>
>   1.
>       When saying:
>       "Please submit the changes as patches, instead of the whole files".  
>       Do you mean to send each file seperately? or other issues as well?
>   2.
>       The change was test and design for 8.1.14, but as far as I see
>       it is also true for any other version.
>       Of course merge is needed in case that the files were changed
>       since then, however , due to the fact that it is only few rows
>       it will be easy to do it.
>   3.
>       Alvaro wrote:
>       "it may be useful to lose the redirection only on the
>        cases where it fails, so we still have reasonable behavior on non-broken
>        platforms"
>
>       Nice idea, but it is really doesn't matter:  on other platform
>       than win32, the code left unchanged! the fix is only relevant
>       for win32 on which (as I explained) the MSI installer (if used)
>       redirect the output by default to a log file.
>
> Regards
> Dror
> ------------------------------------------------------------------------
>
> > Date: Mon, 14 Aug 2006 12:40:25 -0400
> > From: [e-mail]
> > To: [e-mail]
> > CC: [e-mail]
> > Subject: Re: [PATCHES] [Patch] - Fix for bug #2558, InitDB failed to
> run on windows 2003
> >
> > dror wrote:
> >
> > > There were two options to solve this issue:
> > >
> > > Create a new file , grant a write permission for the Postgres user
> > > and redirect the output to that file. (EnterpriseDB  use this method)
> > > Canceling the redirection at all.
> > >  
> > > I choose the second option and omit the redirection in any case that
> > > it windows machine and the redirection was sent to DEVNULL.
> > >  
> > > The only files that I changed are: initDB.c, exec.c and pg_ctl.c
> >
> > Please submit the changes as patches, instead of the whole files.  Also,
> > please specify which branch do these patches apply -- is this for 8.1,
> > or for the current development code?  When checked against the 8.1
> > pg_ctl.c, the file you sent only contains a regression for a bug fix,
> > and no change related to what you describe above.
> >
> > On the other hand, it may be useful to lose the redirection only on the
> > cases where it fails, so we still have reasonable behavior on non-broken
> > platforms.  Or maybe there's a better solution.
> >
> > --
> >
> Alvaro Herrera                                http://www.CommandPrompt.com/
> > PostgreSQL Replication, Consulting, Custom Development, 24x7 support
>
>
> ------------------------------------------------------------------------
> Express yourself instantly with Windows Live Messenger! Windows Live
> Messenger!
> <http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=joinmsncom/messenger>


---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

               http://archives.postgresql.org