Truecrypt Native Linux Drivers Fail about truecrypt HOT 23 CLOSED

Hey FreeApophis... Check Bounty Hunter or whatever it's called. I wrote the solution so TrueCrypt compiles clean on Ubuntu 14.04 LTS. Required some mods but it's working great.

http://s15.postimg.org/4xtjakg1n/Screenshot_22.png

https://www.bountysource.com/issues/5645028-truecrypt-native-linux-drivers-fail/developers

from truecrypt.

// ,

Excellent!

Do you mean Kernel Modules?

from truecrypt.

// ,

Also, does the wxWidgets version have to be 2.8?

In the requirement instructions, it says the following:

• wxWidgets 2.8 shared library and header files installed or wxWidgets 2.8 library source code (available at http://www.wxwidgets.org)

from truecrypt.

// ,

I can't seem to get the wxWidgets header path to work, on this.

I installed FUSE, and added the RSA Security Inc. PKCS #11 Cryptographic Token Interface (Cryptoki) 2.20 header files to a directory defined by the environment variable 'PKCS11_INC', and by 'PKG_CONFIG_PATH'.

I'm not sure what I'm doing wrong, but if you provide a Bitcoin address, I can compensate another $10 for an answer that works.

I LOVE Truecrypt.

➜  TrueCrypt git:(master) ✗ sudo apt-get install fuse
[sudo] password for nathan:
Reading package lists... Done
Building dependency tree
Reading state information... Done
fuse is already the newest version.
The following packages will be REMOVED:
  mfc8840dlpr:i386
0 upgraded, 0 newly installed, 1 to remove and 12 not upgraded.
1 not fully installed or removed.
After this operation, 0 B of additional disk space will be used.
Do you want to continue? [Y/n] y
(Reading database ... 471289 files and directories currently installed.)
Removing mfc8840dlpr (1.1.2-1) ...
/var/lib/dpkg/info/mfc8840dlpr.postrm: 3: /var/lib/dpkg/info/mfc8840dlpr.postrm: /etc/init.d/lpd: not found
dpkg: error processing package mfc8840dlpr (--remove):
 subprocess installed post-removal script returned error exit status 127
Errors were encountered while processing:
 mfc8840dlpr
E: Sub-process /usr/bin/dpkg returned an error code (1)
➜  TrueCrypt git:(master) ✗ export PKCS11_INC=/home/nathan/PKCS11_INC
➜  TrueCrypt git:(master) ✗ PKCS11_INC
~PKCS11_INC
➜  ~PKCS11_INC  echo PKCS11_INC
PKCS11_INC
➜  ~PKCS11_INC  echo $PKCS11_INC
/home/nathan/PKCS11_INC
➜  ~PKCS11_INC  export PKG_CONFIG_PATH=$PKG_CONFIG_PATH:/home/nathan/PKCS11_INC
➜  ~PKCS11_INC  cd ../TrueCrypt
➜  TrueCrypt git:(master) ✗ ls
Boot  Build  Common  Core  Crypto  Driver  Format  License.html  License.txt  Main  Makefile  Mount  pkcs11f.h  pkcs11.h  pkcs11t.h  Platform  Readme.md  Readme.txt  Release  Resources  Volume
➜  TrueCrypt git:(master) ✗ make
Package fuse was not found in the pkg-config search path.
Perhaps you should add the directory containing `fuse.pc'
to the PKG_CONFIG_PATH environment variable
No package 'fuse' found
Package fuse was not found in the pkg-config search path.
Perhaps you should add the directory containing `fuse.pc'
to the PKG_CONFIG_PATH environment variable
No package 'fuse' found
Compiling FuseService.cpp
FuseService.cpp:12:18: fatal error: fuse.h: No such file or directory
 #include <fuse.h>
                  ^
compilation terminated.
/home/nathan/TrueCrypt/Build/Include/Makefile.inc:20: recipe for target 'FuseService.o' failed
make[1]: *** [FuseService.o] Error 1
Makefile:269: recipe for target 'all' failed
make: *** [all] Error 2
➜  TrueCrypt git:(master) ✗ ls ../PKCS11_INC
camellia-contribution.doc  pkcs11.h                   pkcs-11v2-20a2.pdf    pkcs-11v2-20a3.h              pkcs-11v2-20a5d1.pdf   pkcs-11v2-20d3.doc     pkcs-11v2-20.doc               pkcs-11v2-20fd.pdf
cryptoki.h                 pkcs11t-consolidated-d1.h  pkcs-11v2-20a3d1.pdf  pkcs-11v2-20a3.pdf            pkcs-11v2-20agost.doc  pkcs-11v2-20d3.zip     pkcs-11v2-20errata-060515.txt  pkcs-11v2-20fd.zip
ct-kip.h                   pkcs11t.h                  pkcs-11v2-20a3d2.pdf  pkcs-11v2-20a3-withdrawn.pdf  pkcs-11v2-20d1.doc     pkcs-11v2-20d4-cb.doc  pkcs-11v2-20errata.old         pkcs-11v2-20ffd.pdf
otp-pkcs11.h               pkcs11t.h.org              pkcs-11v2-20a3d3.h    pkcs-11v2-20a4d1.pdf          pkcs-11v2-20d1.zip     pkcs-11v2-20d4.doc     pkcs-11v2-20errata.txt         pkcs-11v2-20ffd.zip
pkcs11f.h                  pkcs-11v2-20a1.pdf         pkcs-11v2-20a3d3.pdf  pkcs-11v2-20a4d2.pdf          pkcs-11v2-20d2.zip     pkcs-11v2-20d4.zip     pkcs-11v2-20fd.doc             pkcs-11v2-20.pdf
➜  TrueCrypt git:(master) ✗

from truecrypt.

See... this is actually my one complaint about the Bounty Source site... It's a cool idea , novel, lots of potential, and for a good cause... but the execution leaves alot to be desired... ie. "Ideal World" -- We should both have well defined expectations of a software 'bounty' (ie. You have your specific problem defined by constraints A,B,C, I agree to fix it subject to paramters D,E,F, then IF my solution meets your public 'acceptence criteria' YZ, then I expect to be paid amount G within time T.

This was my first experience on the site after reading a review, and I thought 'this is perfect' , since its TrueCrypt, it's a great piece of software, you had the exact same problem I've run into like 3x before, etc. So I'm quite pleased I was able to solve it. . . .I think everyone should have free strong crypto.

Anyway, so this was more of a test (or 'social experiment', for me) to see if the MVP Bounty Source system added value -- perhaps by tracking requirements, or by acting as escrow for both the money and the source code, etc -- without alot of risk for either of us.

But it's clear we'll just have to work directly since those features aren't in the platform. So no worries. On to the solution, what I did, how I did it, what we want to do next, etc.

So let me make try to clarify my side of this exchange..

(1) I made the assumption that you had problems getting Truecrypt to work on Ubuntu 14.04 amd64 ... I surmised you ran into compile issues with wxWidgets etc, or that if you got it compiled, there was some problem with linking into the shared libraries in userspace or their include files . I was familiar with this exact issue since I do a fair amount of crypto work, and the same problem compiling TrueCrypt on Ubuntu for the past 3 years with each LTS release (TrueCrypt is a pain in the neck to get working, and it gets worse the older the code gets).

(2) To answer your earlier question, no, I found no issue with kernel modules, and would be surprised to learn if that's a problem. TrueCrypt's primary use-case is a userspace process as root. I think it can install itself to the MBR , but that's not a common requirement. Anyway -- My goal was to get it working (compiled, linked, running) on both the command line and in an X desktop environment like Mate or XFCE.

(3) I started by sorting out which wx packages were required , which needed to be purged, and which were optional . Some wx packages are required for command line use. Others are required for the GUI to work I got this identified and separated. This involved purging a couple of unwanted wx libs using apt-get purge. Yes, you are correct -- TrueCrypt wants wx2.8 . I think it may work with later version, but 2.8 is ideal , and that's what I used.

(4) I used all built-in Ubuntu packages... No extra universe repos or anything like that . Just vanilla Ubuntu packages, but a specific subset of the wx2.8 ones. There are a few other libraries you need like apt-get install nasm , a couple of others , nothing unusual though.

(5) I inadvertently read your post that perhaps the issue maybe had to do with lsb_release. I patched this script to solve an open Ubuntu issue that affects at least one application, and I have a patch file. I don't know whether this was part of the issue for TrueCrypt. I think it was unrelated in hindsight.

(6) Next, Truecrypt would compile most of the way after fixing the apt-get packages , but would fail on some information related to wx. Eventually I realized it was not an issue in the linker, but rather in the header files. For some reason (on my box) , the include files were offset up one directory. I solved this by (1) a symlink, and then (2) manually copying a directory of wx .h files found in /usr/include/wx to exist in the path /usr/include/wx-2.8/wx (where truecrypt was looking for them).

(7) Now the compiler errors dropped dramatically. The final issues involved in realizing that TrueCrypt actually depends on Qt4. This package was installed already via apt, but again the issue was with pointing to the wrong header path. I modified the Qt4 files in two ways... (1) I encased all of them with #ifndef __FILENAME_H #define __FILENAME_H /* header file data goes here */ #endif ... This prevented circular dependencies where header1 imports header2 imports header1 etc. Then (2) I modified the #include paths to change the paths from #include <QtCore/whatever.h> to <qt4/QtCore/whatever.h>, thus matching the on-disk layout in /usr/include.

(8) I re-ordered the import of the QtCore header files in Truecrypt's main include files so that the correct definitions (ie QChar) were imported first.

(9) Eventually I reduced the compile to two errors...

Compiling CommandLineInterface.cpp
Application.h: In constructor ‘TrueCrypt::CommandLineInterface::CommandLineInterface(wxCmdLineParser&, TrueCrypt::UserInterfaceType::Enum)’:
Application.h:409:46: error: ‘class wxString’ has no member named ‘ToStdWstring’
Application.h:500:58: error: ‘class wxString’ has no member named ‘ToStdWstring’
/chome/user/tc/TrueCrypt/Build/Include/Makefile.inc:20: recipe for target 'CommandLineInterface.o' failed
make[1]: *** [CommandLineInterface.o] Error 1
Makefile:269: recipe for target 'all' failed
make: *** [all] Error 2

(10) This was the hardest part since I didn't remember how I fixed it. Finally I realized that it had to do with the way Qt4 was compiled for Ubuntu 14 amd64... A certain flag was not properly set, slightly reducing the width of some of the wxStrings. I believe this is mainly of no functional impact since it's isolated and GUI specific to two lines of code. My workaround solution was either (1) download and recompile Qt4, or (2) change the TrueCrypt source to use wxString::toAscii() rather than wxString::toStrWstring() (the latter was not properly compiled into my Qt4 based on the wxUSE_STL flag).

So after I reverted to the toAscii() functions , I got a clean compile (no errors, a fair amount of warnings but no functional impact). I then verified TrueCrypt with a few test vectors and it is working properly.

Next Steps:
(1) I need from you exactly what goes wrong when you were trying to build. You can cut and paste and email it to me if you want .. that way I can check my work.

(2) I will test the solution in a fresh ubuntu 14.04 and 12.04 virtual machines to make sure TC works and is interoperable with all other recent TC versions.

(3) I'll then either send you a .tgz file + README, or give you the .vmdks , whatever you prefer, and you can verify the solution. Alternately I can commit it here if it's just code and not a vm or container.

Okay that's all for now. Let me know how you want me to share the work, and I'll double check it in virtualbox, then share it.

from truecrypt.

Hold on, a moment, I think I may have screwed this up. Just give me another
20 minutes, here, and I think I can do it.

On Sun, Apr 19, 2015 at 12:29 AM, Alex [email protected] wrote:

See... this is actually my complaint about the Bounty Source site... It's
a cool idea , but the execution leaves alot to be desired... ie. We should
both have well defined expectations (ie. You have problem defined by
constraints A,B,C, I agree to fix it subject to paramters D,E,F, then if
the solution meets public 'acceptence criteria', then I expect to be paid
amount G within time T.

Anyway so this was more of a test of the MVP Bounty Source system on a
simple problem -- without alot of risk for either of us. On to the
solution...

So let me make try to clarify my side of this exchange..

(1) I made the assumption that you had problems getting Truecrypt to work
on Ubuntu 14.04 amd64 ... I surmised you ran into compile issues with
wxWidgets etc, or that if you got it compiled, there was some problem with
linking into the shared libraries in userspace or their include files . I
was familiar with this exact issue since I do a fair amount of crypto work,
and the same problem compiling TrueCrypt on Ubuntu for the past 3 years
with each LTS release (TrueCrypt is a pain in the neck to get working, and
it gets worse the older the code gets).

(2) To answer your earlier question, no, I found no issue with kernel
modules, and would be surprised to learn if that's a problem. TrueCrypt's
primary use-case is a userspace process as root. I think it can install
itself to the MBR , but that's not a common requirement. Anyway -- My goal
was to get it working (compiled, linked, running) on both the command line
and in an X desktop environment like Mate or XFCE.

(3) I started by sorting out which wx packages were required , which
needed to be purged, and which were optional . Some wx packages are
required for command line use. Others are required for the GUI to work I
got this identified and separated. This involved purging a couple of
unwanted wx libs using apt-get purge. Yes, you are correct -- TrueCrypt
wants wx2.8 . I think it may work with later version, but 2.8 is ideal ,
and that's what I used.

(4) I used all built-in Ubuntu packages... No extra universe repos or
anything like that . Just vanilla Ubuntu packages, but a specific subset of
the wx2.8 ones. There are a few other libraries you need like apt-get
install nasm , a couple of others , nothing unusual though.

(5) I inadvertently read your post that perhaps the issue maybe had to do
with lsb_release. I patched this script to solve an open Ubuntu issue that
affects at least one application, and I have a patch file. I don't know
whether this was part of the issue for TrueCrypt. I think it was unrelated
in hindsight.

(6) Next, Truecrypt would compile most of the way after fixing the apt-get
packages , but would fail on some information related to wx. Eventually I
realized it was not an issue in the linker, but rather in the header files.
For some reason (on my box) , the include files were offset up one
directory. I solved this by (1) a symlink, and then (2) manually copying a
directory of wx .h files found in /usr/include/wx to exist in the path
/usr/include/wx-2.8/wx (where truecrypt was looking for them).

(7) Now the compiler errors dropped dramatically. The final issues
involved in realizing that TrueCrypt actually depends on Qt4. This package
was installed already via apt, but again the issue was with pointing to the
wrong header path. I modified the Qt4 files in two ways... (1) I encased
all of them with #ifndef __FILENAME_H #define __FILENAME_H /* header file
data goes here */ #endif ... This prevented circular dependencies where
header1 imports header2 imports header1 etc. Then (2) I modified the
#include paths to change the paths from #include to , thus matching the
on-disk layout in /usr/include.

(8) I re-ordered the import of the QtCore header files in Truecrypt's main
include files so that the correct definitions (ie QChar) were imported
first.

(9) Eventually I reduced the compile to two errors...

Compiling CommandLineInterface.cpp
Application.h: In constructor
‘TrueCrypt::CommandLineInterface::CommandLineInterface(wxCmdLineParser&,
TrueCrypt::UserInterfaceType::Enum)’:
Application.h:409:46: error: ‘class wxString’ has no member named
‘ToStdWstring’
Application.h:500:58: error: ‘class wxString’ has no member named
‘ToStdWstring’
/chome/user/tc/TrueCrypt/Build/Include/Makefile.inc:20: recipe for target
'CommandLineInterface.o' failed
make[1]: *** [CommandLineInterface.o] Error 1
Makefile:269: recipe for target 'all' failed
make: *** [all] Error 2

(10) This was the hardest part since I didn't remember how I fixed it.
Finally I realized that it had to do with the way Qt4 was compiled for
Ubuntu 14 amd64... A certain flag was not properly set, slightly reducing
the width of some of the wxStrings. I believe this is mainly of no
functional impact since it's isolated and GUI specific to two lines of
code. My workaround solution was either (1) download and recompile Qt4, or
(2) change the TrueCrypt source to use wxString::toAscii() rather than
wxString::toStrWstring() (the latter was not properly compiled into my Qt4
based on the wxUSE_STL flag).

So after I reverted to the toAscii() functions , I got a clean compile (no
errors, a fair amount of warnings but no functional impact). I then
verified TrueCrypt with a few test vectors and it is working properly.

Next Steps:
(1) I need from you exactly what goes wrong when you were trying to build.
You can cut and paste and email it to me if you want .. that way I can
check my work.

(2) I will test the solution in a fresh ubuntu 14.04 and 12.04 virtual
machines to make sure TC works and is interoperable with all other recent
TC versions.

(3) I'll then either send you a .tgz file + README, or give you the .vmdks
, whatever you prefer, and you can verify the solution. Alternately I can
commit it here if it's just code and not a vm or container.

Okay that's all for now. Let me know how you want me to share the work,
and I'll double check it in virtualbox, then share it.

—
Reply to this email directly or view it on GitHub
#18 (comment)
.

from truecrypt.

I think I may have made a mistake while compiling FUSE.

On Sun, Apr 19, 2015 at 12:32 AM, Nate B [email protected] wrote:

Hold on, a moment, I think I may have screwed this up. Just give me
another 20 minutes, here, and I think I can do it.

On Sun, Apr 19, 2015 at 12:29 AM, Alex [email protected] wrote:

See... this is actually my complaint about the Bounty Source site... It's
a cool idea , but the execution leaves alot to be desired... ie. We should
both have well defined expectations (ie. You have problem defined by
constraints A,B,C, I agree to fix it subject to paramters D,E,F, then if
the solution meets public 'acceptence criteria', then I expect to be paid
amount G within time T.

Anyway so this was more of a test of the MVP Bounty Source system on a
simple problem -- without alot of risk for either of us. On to the
solution...

So let me make try to clarify my side of this exchange..

(1) I made the assumption that you had problems getting Truecrypt to work
on Ubuntu 14.04 amd64 ... I surmised you ran into compile issues with
wxWidgets etc, or that if you got it compiled, there was some problem with
linking into the shared libraries in userspace or their include files . I
was familiar with this exact issue since I do a fair amount of crypto work,
and the same problem compiling TrueCrypt on Ubuntu for the past 3 years
with each LTS release (TrueCrypt is a pain in the neck to get working, and
it gets worse the older the code gets).

(2) To answer your earlier question, no, I found no issue with kernel
modules, and would be surprised to learn if that's a problem. TrueCrypt's
primary use-case is a userspace process as root. I think it can install
itself to the MBR , but that's not a common requirement. Anyway -- My goal
was to get it working (compiled, linked, running) on both the command line
and in an X desktop environment like Mate or XFCE.

(3) I started by sorting out which wx packages were required , which
needed to be purged, and which were optional . Some wx packages are
required for command line use. Others are required for the GUI to work I
got this identified and separated. This involved purging a couple of
unwanted wx libs using apt-get purge. Yes, you are correct -- TrueCrypt
wants wx2.8 . I think it may work with later version, but 2.8 is ideal ,
and that's what I used.

(4) I used all built-in Ubuntu packages... No extra universe repos or
anything like that . Just vanilla Ubuntu packages, but a specific subset of
the wx2.8 ones. There are a few other libraries you need like apt-get
install nasm , a couple of others , nothing unusual though.

(5) I inadvertently read your post that perhaps the issue maybe had to do
with lsb_release. I patched this script to solve an open Ubuntu issue that
affects at least one application, and I have a patch file. I don't know
whether this was part of the issue for TrueCrypt. I think it was unrelated
in hindsight.

(6) Next, Truecrypt would compile most of the way after fixing the
apt-get packages , but would fail on some information related to wx.
Eventually I realized it was not an issue in the linker, but rather in the
header files. For some reason (on my box) , the include files were offset
up one directory. I solved this by (1) a symlink, and then (2) manually
copying a directory of wx .h files found in /usr/include/wx to exist in the
path /usr/include/wx-2.8/wx (where truecrypt was looking for them).

(7) Now the compiler errors dropped dramatically. The final issues
involved in realizing that TrueCrypt actually depends on Qt4. This package
was installed already via apt, but again the issue was with pointing to the
wrong header path. I modified the Qt4 files in two ways... (1) I encased
all of them with #ifndef __FILENAME_H #define __FILENAME_H /* header file
data goes here */ #endif ... This prevented circular dependencies where
header1 imports header2 imports header1 etc. Then (2) I modified the
#include paths to change the paths from #include to , thus matching the
on-disk layout in /usr/include.

(8) I re-ordered the import of the QtCore header files in Truecrypt's
main include files so that the correct definitions (ie QChar) were imported
first.

(9) Eventually I reduced the compile to two errors...

Compiling CommandLineInterface.cpp
Application.h: In constructor
‘TrueCrypt::CommandLineInterface::CommandLineInterface(wxCmdLineParser&,
TrueCrypt::UserInterfaceType::Enum)’:
Application.h:409:46: error: ‘class wxString’ has no member named
‘ToStdWstring’
Application.h:500:58: error: ‘class wxString’ has no member named
‘ToStdWstring’
/chome/user/tc/TrueCrypt/Build/Include/Makefile.inc:20: recipe for target
'CommandLineInterface.o' failed
make[1]: *** [CommandLineInterface.o] Error 1
Makefile:269: recipe for target 'all' failed
make: *** [all] Error 2

(10) This was the hardest part since I didn't remember how I fixed it.
Finally I realized that it had to do with the way Qt4 was compiled for
Ubuntu 14 amd64... A certain flag was not properly set, slightly reducing
the width of some of the wxStrings. I believe this is mainly of no
functional impact since it's isolated and GUI specific to two lines of
code. My workaround solution was either (1) download and recompile Qt4, or
(2) change the TrueCrypt source to use wxString::toAscii() rather than
wxString::toStrWstring() (the latter was not properly compiled into my Qt4
based on the wxUSE_STL flag).

So after I reverted to the toAscii() functions , I got a clean compile
(no errors, a fair amount of warnings but no functional impact). I then
verified TrueCrypt with a few test vectors and it is working properly.

Next Steps:
(1) I need from you exactly what goes wrong when you were trying to
build. You can cut and paste and email it to me if you want .. that way I
can check my work.

(2) I will test the solution in a fresh ubuntu 14.04 and 12.04 virtual
machines to make sure TC works and is interoperable with all other recent
TC versions.

(3) I'll then either send you a .tgz file + README, or give you the
.vmdks , whatever you prefer, and you can verify the solution. Alternately
I can commit it here if it's just code and not a vm or container.

Okay that's all for now. Let me know how you want me to share the work,
and I'll double check it in virtualbox, then share it.

—
Reply to this email directly or view it on GitHub
#18 (comment)
.

from truecrypt.

I'll put all this in a README if you want,so you don't have to replicate this from what I've written. It's tedious. I only did it out of respect for TrueCrypt . *pours out some beer

I'll give you the code and README tomorrow , once I sleep and double-check my work. Then you can check on your machine, and you can also replicate the fix in the future from the rough documentation I've written on how I fixed it. Last note: I used the TrueCrypt code in your repo, whatever that version is.

Ah yeah FUSE is picky as hell. Try this. I hope you have python 2.6/2.7.

http://www.stavros.io/posts/python-fuse-filesystem/

Cut and paste that tutorial. Run the main file as python ./fuse_fs.py /mydir /mnt/fuse

Dump files into /mydir ... they should appear in /mnt/fuse magically.. Delete them from /mnt/fuse.. . they should disappear from /mydir instantly. Last, do du -h and ls -al on both directory contents. If fuse is working, you should have all that working, and there will only be one actual disk copy of the files (just two 'views' of them).

If that python script works, then fuse is probably ok on your box.

from truecrypt.

Ah. I have been checking this out from FreeApophis/TrueCrypt to my local,
for testing, and keeping the older code up on my account for later
comparison.

On Sun, Apr 19, 2015 at 12:43 AM, Alex [email protected] wrote:

And I'll put all this in a README so don't worry about trying to replicate
this from what I've written. I'll give you the code and README tomorrow ,
once I sleep and double-check my work. Then you can check on your machine,
and you can also replicate the fix in the future from the rough
documentation I've written on how I fixed it. Last note: I used the
TrueCrypt code in your repo, whatever that version is.

—
Reply to this email directly or view it on GitHub
#18 (comment)
.

from truecrypt.

I seem to have run into a (6) from your previous message, with wx's path being offset by a directory:

➜  TrueCrypt git:(master) ✗ make
Compiling FuseService.cpp
Updating library Driver.a
Compiling CoreBase.cpp
Compiling CoreException.cpp
Compiling FatFormatter.cpp
Compiling HostDevice.cpp
Compiling MountOptions.cpp
Compiling RandomNumberGenerator.cpp
Compiling VolumeCreator.cpp
Compiling CoreService.cpp
Compiling CoreServiceRequest.cpp
Compiling CoreServiceResponse.cpp
Compiling CoreUnix.cpp
Compiling CoreLinux.cpp
Updating library Core.a
make[1]: wx-config: Command not found
make[1]: wx-config: Command not found
Precompiling SystemPrecompiled.h
SystemPrecompiled.h:9:19: fatal error: wx/wx.h: No such file or directory
 #include <wx/wx.h>
                   ^
compilation terminated.
/home/nathan/TrueCrypt/Build/Include/Makefile.inc:30: recipe for target 'SystemPrecompiled.h.gch' failed
make[1]: *** [SystemPrecompiled.h.gch] Error 1
Makefile:269: recipe for target 'all' failed
make: *** [all] Error 2

From the earlier message:

I seem to have run into (6):

(6) Next, Truecrypt would compile most of the way after fixing the apt-get packages , but would fail on some information related to wx. Eventually I realized it was not an issue in the linker, but rather in the header files. For some reason (on my box) , the include files were offset up one directory. I solved this by (1) a symlink, and then (2) manually copying a directory of wx .h files found in /usr/include/wx to exist in the path /usr/include/wx-2.8/wx (where truecrypt was looking for them).

from truecrypt.

Okay, that's fine I think... right? I used FreeApophis/TrueCrypt as well. I forked your repo to my github as bitshark ,then cloned it from there to my laptop at home..

user@computer:~/tc/TrueCrypt$ git branch -a

• master
  remotes/origin/7.2-warn
  remotes/origin/HEAD -> origin/master
  remotes/origin/master

Below is the last commit on master branch before I did all that #include sorcery to deal with Ubuntu Corp's tendency to move things around for no reason

commit bc57fd6
Merge: 0ccbbe4 de41442
Author: Thomas [email protected]
Date: Tue Jun 10 00:42:02 2014 +0200

Merge pull request #8 from pmalek/patch-1

Update Readme.md with code styling.

from truecrypt.

Yeah, that's actually good news. That sounds identical. Lemme see if I can put in a pull request right quick . This /usr/include hack is not pretty, so prepare yourself haha

from truecrypt.

My body is ready. Keep in mind that I have much more experience at compiling Java, Elixir, and Python than anything C related.

from truecrypt.

OK. Fixed the fuse compilation error, but I'm still getting an issue with wx. I can't seem to get anything wx related to show up in /usr/include/, and although I've set the environment variables properly.

from truecrypt.

I'm now trying to put the wx files in a standard include path. Perhaps this is what you meant by the folders related to wx-2.8, above?

➜  ~  sudo cp -r PKCS11_INC/wx /usr/include/
➜  ~  sudo mkdir /usr/include/wx-2.8
➜  ~  sudo cp -r PKCS11_INC/wx /usr/include/wx-2.8
➜  ~  ls /usr/include/wx-2.8 
wx
➜  ~  

I still achieve only the following:

➜  TrueCrypt git:(master) ✗ make
make[1]: wx-config: Command not found
make[1]: wx-config: Command not found
Precompiling SystemPrecompiled.h
SystemPrecompiled.h:9:19: fatal error: wx/wx.h: No such file or directory
 #include <wx/wx.h>
                   ^
compilation terminated.
/home/nathan/TrueCrypt/Build/Include/Makefile.inc:30: recipe for target 'SystemPrecompiled.h.gch' failed
make[1]: *** [SystemPrecompiled.h.gch] Error 1
Makefile:269: recipe for target 'all' failed
make: *** [all] Error 2

from truecrypt.

Okay haha.. .word allright so I did this like lazymode ... I put the /usr/include hax and all the source code hax into git. I also put in a quick snapshot of what my dpkg -l wx lookz like on a testbed... And lastly I've written some real quick readme's on what the heck I did to complement the comments.

This might be sufficient to get it running ... if not , let me know .... next step we'll test on an lxc container or on one of my amis -- or a vps or sumt -- only if this becomes difficult. I think it'll work tho.

from truecrypt.

Ok, pull request is open! I'm gonna go wind down, but good luck with this . Let's catch up tomorrow and see if it works. Don't spend like 6h on it... If it doesn't work within like 90min, let's just both log into an amazon ami over ssh and test there tomrorow. Itz been fun... good luck buddy! Hopefully you'll have AES-Serpent-Twofish in like 20 mins

Ah okay, I think we're gonna need to compare your dpkg -l to my dpkg -l. I included a partial one . . . We can compare the entire dpkg -l cross reference tmo if its not working.

from truecrypt.

// , You are amazing.

This is precisely what I would have asked for, especially the snapshot of what your dpkg -l wx looks like on a testbed.

➜ ~PKCS11_INC sudo apt-get install wx2.8-headers
[sudo] password for nathan:
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following NEW packages will be installed:
wx2.8-headers
0 upgraded, 1 newly installed, 1 to remove and 12 not upgraded.
1 not fully installed or removed.
Need to get 0 B/1,022 kB of archives.
After this operation, 6,988 kB of additional disk space will be used.
Do you want to continue? [Y/n] y
(Reading database ... 471289 files and directories currently installed.)

Apparently I need to do this on a VM. I keep getting the following error:

➜ ~PKCS11_INC sudo apt-get install wx2.8-headers
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following packages will be REMOVED:
mfc8840dlpr:i386
The following NEW packages will be installed:
wx2.8-headers
0 upgraded, 1 newly installed, 1 to remove and 12 not upgraded.
1 not fully installed or removed.
Need to get 0 B/1,022 kB of archives.
After this operation, 6,988 kB of additional disk space will be used.
Do you want to continue? [Y/n] y
(Reading database ... 471289 files and directories currently installed.)
Removing mfc8840dlpr (1.1.2-1) ...
/var/lib/dpkg/info/mfc8840dlpr.postrm: 3: /var/lib/dpkg/info/mfc8840dlpr.postrm: /etc/init.d/lpd: not found
dpkg: error processing package mfc8840dlpr (--remove):
subprocess installed post-removal script returned error exit status 127
Errors were encountered while processing:
mfc8840dlpr
E: Sub-process /usr/bin/dpkg returned an error code (1)

from truecrypt.

I merged the pull-request in its own branch -> native_linux:

This repo is only a history upto 7.1a with all commits, I try to keep it clean to a bare minimum. Development on this base has progressed with CipherShed and VeraCrypt already quite far away from this repository.

We as, pure-privacy.org support the Projects CipherShed and VeraCrypt as the future replacements of TrueCrypt.

from truecrypt.

// , I'd like to think that there still is hope:

https://www.grc.com/misc/truecrypt/truecrypt.htm

from truecrypt.

Excellent! I'm glad to hear it, Thomas.

Why is it, then, that https://truecrypt.ch/ links to FreeApophis/TrueCrypt
under the "Hack some code" label?

Should it not link to something like
https://veracrypt.codeplex.com/SourceControl/network?

Have Alex and I wasted our time?

On Sun, Apr 19, 2015 at 1:24 AM, Thomas [email protected] wrote:

I merged the pull-request in its own branch -> native_linux:

This repo is only a history upto 7.1a with all commits, I try to keep it
clean to a bare minimum. Development on this base has progressed with
CipherShed and VeraCrypt already quite far away from this repository.

We as, pure-privacy.org supports the Projects CipherShed and VeraCrypt as
the future replacements of TrueCrypt.

—
Reply to this email directly or view it on GitHub
#18 (comment)
.

from truecrypt.

Currently no one is activly working on this code directly. We would need a new maintainer to do that. I am not deep enough in the code at the moment to do this, and 2 other projects try 2 different ways to keep truecrypt running - so this attempt which was one of the first went into a bit of silent mode. And there is the Legal Problem which is discussed in some other tickets about the licensing of this code especially the name.

If you want to get in contact with the pure-privacy team or either Projects feel free to contact us directly. [email protected]

You certainly have not wasted your time, you solved your problem, and it is now available in the main repository. Thank you very much for your effort. I will also merge further improvements.

I also added you both to the collaborators of this projects.

from truecrypt.

The development and support of TrueCrypt has ended,

The further Development supported by https://pure-privacy.org/ has moved to:
https://github.com/CipherShed/CipherShed/issues

As an alternative, you can also use the backwards compatibility of VeraCrypt:
https://github.com/veracrypt/VeraCrypt

from truecrypt.