Jump to content


 


Register a free account to unlock additional features at BleepingComputer.com
Welcome to BleepingComputer, a free community where people like yourself come together to discuss and learn how to use their computers. Using the site is easy and fun. As a guest, you can browse and view the various discussions in the forums, but can not create a new topic or reply to an existing one unless you are logged in. Other benefits of registering an account are subscribing to topics and forums, creating a blog, and having no ads shown anywhere on the site.


Click here to Register a free account now! or read our Welcome Guide to learn how to use this site.

Photo

The library boost bootstrap.bat leads to Access is Denied errors.


  • Please log in to reply
10 replies to this topic

#1 PetarSickey

PetarSickey

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 10:40 AM

I had originally wanted to post this on a more programming oriented forum but the forum is not working properly right now when I try to post, so
I've brought the problem here.  It is pretty much a general Windows 10 question anyway.

 

I am running bootstrap.bat on my Windows 10 machine which has the library called boost installed (boost 5.6)

The batch file bootstrap.bat, which in turn calls b2.exe and bjam.exe, gives the error "Access is denied". 

However even if I run the command prompt window as "an Administrator", I still get the same error: "Access is Denied".

On the same topic, does the chmod command work in windows.  I know I can change security in the properties tab
for directories and files, but when I look in those directories in Windows 10, they all have at least read and execute priveleges.  Even
for non-administrators.  So b2.exe and bjam.exe should execute.  They shouldn't need me to change the permissions.

I didn't have this problem on my other Windows 10 machine, and it was at least only occasionally that I had to run something as
Administrator.  I probably took the files for boost from that machine (which is now running FatDog64, an OS which creates a

virtual file layer).  Could this be the problem?  Perhaps I need to do yet another download - this time a fresh copy of boost?  Just

a thought that just occurred to me now.

Any clue why I cannot run bootstrap without this error or what I should try now?  Should I just redownload boost and rerun the command

that's troubling me.



BC AdBot (Login to Remove)

 


#2 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 12:20 PM

I don't know maybe it's a 7zip issue?



#3 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 12:30 PM

I'm downloading the .zip version.  Before, I used 7z and it worked fine, but that's a different machine.  Perhaps my new installation of .7z was not good enough for some reason - I DO remember a lot of permission problems.  ??

 

The boost 1.5.6 is only 154 mB zipped - not that long a download.  I'll have to reinstall.



#4 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 02:53 PM

I just unzipped boost 1.5.6.0, the same version as before - why is there no b2.exe or bjam.exe?  Why is it different than before?

Just curious, perhaps it's due to some process.

Anyway, I just ran bootstrap and I did NOT seem to have any troubles with it.  It ran to completion with no strange errors, so I can proceed with

.\b2 from here.

 

I guess trying a different zip was a good idea.  I'll have to hunt around and try to understand the bootstrap better, and then explain why no b2/bjam.

Perhaps it changed.


Edited by PetarSickey, 25 August 2016 - 02:57 PM.


#5 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 03:03 PM

So I'm running .\b2 now ..., just the ususal conversion of type warnings so far, no errors ... fingers crossed.

 

Epilogue: boost has built identically to before: failed to update 5 targets, skipped 12 targets, updated 1198 targets.  So things are ok now.  Probably ok to close this thread - the cause was "something freaky" with either a .7z file or the process I used to extract.  I don't know what the exact problem was (at least yet).


Edited by PetarSickey, 25 August 2016 - 04:08 PM.


#6 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 03:13 PM

I deleted the original boost directory and started over and it seems ok.  I suspect it's either the .7z I used, or perhaps the program I installed to use .7z files for extracting and archiving, or possibly both.


Edited by PetarSickey, 25 August 2016 - 03:14 PM.


#7 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 04:49 PM

I now realize that boost is missing a couple libs having to do with wserial and zlib.  ?  Someone changed the build of boost (with 1.56.0), so that now

building zlib is an option (and in my output near the beginning, it says - zlib                         : no    (cached) whereas in the original build there was no

mention of zlib since it was assumed build I guess).  So I need to make it build zlib.  Not sure about wserial.

 

I wonder if I need to remove a space before the path, eg.   -sZLIB_SOURCE= rather than -s ZLIB_SOURCE=.  I had a space originally.


Edited by PetarSickey, 25 August 2016 - 05:08 PM.


#8 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 05:18 PM

Wow, I now remember that it's just a file is missing from the build - no big deal. 

I found my old notes:

 

"When I try to set the BOOST_LIBRARYDIR or BOOST_ROOT to the stage\lib it won't let me - complains of error in project file.
Perhaps my ZLIB_SOURCE was not correct - I need to investigate this. I know what directory the zlib stuff is in - I gave it a path
above that - I assume it would recurse down - bad assumption?
There is a file missing from the build called unistd.h and this is why the libs for zlib are not built. I have to find that file.
I can get zlib to build by editing the zconf.h since Unix build not needed but perhaps I need to look further into Python not being set
up and it's ability to detect which compiler I have. Perhaps I can try to Configure cmake now, perhaps not."

 

Sure glad I had kept some notes from June 1 of this year.  So tonight or tomorrow I can fix this problem probably the same way I did before.

:tophat:

Wow, I now realize that I already had made the changes to zconf.h (I just copied my old zlib directory).  So it's something else.


Edited by PetarSickey, 25 August 2016 - 05:28 PM.


#9 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 25 August 2016 - 09:17 PM

I need to understand .jam files.  Tomorrow.



#10 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 26 August 2016 - 06:58 AM

Under bin.v2, I see a project-cache.jam that always says "missing" next to the line that builds zlib, even when it DOES build it.  And my new project has not binary directory zlib under standalone which is obvious since zlib is not building.  My old system had that binary directory.  I need to get zlib to build as it has iostream.  I need to find why it does not build despite my including.  It doesn't seem like an iterative build issue.  It's just not included.



#11 PetarSickey

PetarSickey
  • Topic Starter

  • Members
  • 151 posts
  • OFFLINE
  •  
  • Local time:04:46 AM

Posted 31 August 2016 - 06:14 AM

I almost forgot to wrap up this thread.

 

I solved all the boost/zlib build problems.  The problem was simply that there was one build of boost and zlib that had an error in an iostream file (I believe it was an iostream file).

This error caused the library to fail to build the zlib (2 libs, 2 dlls).  I simply used a different (later) version of zlib, and both problems went away.   Naturally, any error we see at

the object level would prevent a final link of the two libs.  It was just a minor error in a source file.  Perhaps it was overlooked in that version of zlib.  Perhaps it built on some people's machines for some reason.  On my machine, a later version of zlib was the pill needed for the cure.

 

So the build goes fine without any tweaking now.  One can simply use the ZLIB_SOURCE option and provide the path.

So this thread can be closed forever, unless any comments or questions.

 

Come to think of it, it might have been boost_log_vc140-mt-gd-1_56.dll that was not building (unresolved external in unhandled_exception_count.obj).

Anyway, a different zlib fixed the problem.  I had thought it was something with iostream.  The boost_zlib-vc140-.....lib/dll were failing to build.  So both

these two sets of libs had failed to build.  You'll note that they both have the -mt-gd-1_56.dll on the end for the debug.

 

So the log is under lib/logs and the zlib is under standalone.  Both had failed to build for some weird reason.


Edited by PetarSickey, 31 August 2016 - 06:23 AM.





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users