Corrupt .pdb file (error #31000)

Corrupt .pdb file (error #31000)

Аватар пользователя Alex K.

Hi again,

Hopefully it's not too annoying me posting these threads here, I've been trying for a few months unsuccessfully to compile Firefox with ICC x64 and feel like I've made quite a lot of progress but I'm not quite there yet.

This error is strange because I can hardly find any information on it and very few people seem to get it.

The error

I've attached the relevant files. I've made sure that the source files know to use ICL as opposed to CL (one of my previous mistakes) and this error occurs. It happens on Firefox 23 and Firefox 24 Beta (any version). Is this usually caused by user error?

I've tried using -Z7 but somewhere a makefile is overwriting it with -Zi

ВложениеРазмер
Скачать snip2.png73.46 КБ
Скачать log.txt230.91 КБ
Скачать debug.i79.33 КБ
Скачать debug.c2.18 КБ
15 сообщений / 0 новое
Последнее сообщение
Пожалуйста, обратитесь к странице Уведомление об оптимизации для более подробной информации относительно производительности и оптимизации в программных продуктах компании Intel.
Аватар пользователя Sergey Kostrov

>>...I've tried using -Z7 but somewhere a makefile is overwriting it with -Zi...

Alex, Please take a look at modified screenshot:

Вложения: 

ВложениеРазмер
Скачать snip2.png82.69 КБ
Аватар пользователя Melanie Blower (Intel)

Your screenshot contains a message that the filesystem is out of space. This could also cause debug file corruption.

The Intel compiler recently added support for a system environment variable that pre- or post- pends compiler flags onto each compilation line. It could be useful in your case where there's a complicated build system and you're not sure how to modify all the /Zi flags into /Z7

You could set the environment variable __INTEL_POST_CFLAGS to be -Z7. Since Z7 occurs later on the command line, it will override any Zi that may occur earlier in the command line.

Аватар пользователя Sergey Kostrov

>>...You could set the environment variable __INTEL_POST_CFLAGS...

Is it described in documentation?

Аватар пользователя Melanie Blower (Intel)
Аватар пользователя Sergey Kostrov

>>It's in these release notes:
>>
>>http://software.intel.com/sites/default/files/article/251095/release-not...

My question is:

Why Release Notes are Not updated on http://software.intel.com/en-us/articles/intel-parallel-studio-xe-2013-r... web-page?

Аватар пользователя Sergey Kostrov

>>...Hopefully it's not too annoying me posting these threads here, I've been trying for a few months unsuccessfully to
>>compile Firefox with ICC x64 and feel like I've made quite a lot of progress but I'm not quite there yet...

I assume that you're doing this just for yourself, right? Or, Is it a Firefox Community related task?

Аватар пользователя Sergey Kostrov

>>...Why Release Notes are Not updated on http://software.intel.com/en-us/articles/intel-parallel-studio-xe-2013-r... web-page?

If you look at a pdf file ( any Update for Windows ) you could find the following:

...
1.2 Product Contents
Intel(R) Parallel Studio XE 2013 includes the following components:
- Intel(R) C++ Composer XE 2013 Update 5 - includes Intel(R) Integrated Performance Primitives (Intel(R) IPP), Intel(R) Threading Building Blocks (Intel(R) TBB) and Intel(R) Math Kernel Library (Intel(R) MKL)
...

Q: What are versions and updates for IPP, TBB and MKL libraries?

Аватар пользователя bob j.

Цитата:

Sergey Kostrov wrote:

>>...Hopefully it's not too annoying me posting these threads here, I've been trying for a few months unsuccessfully to
>>compile Firefox with ICC x64 and feel like I've made quite a lot of progress but I'm not quite there yet...

I assume that you're doing this just for yourself, right? Or, Is it a Firefox Community related task?

thank for helping alex cant wait for new release of waterfox Waterfox | Downloads | The fastest 64-Bit variant of Firefox!
he has had loots of issues build waterfox for his user base please help him were all waiting for a new release none since watefox 18 please help him thank :)

Аватар пользователя bob j.

cant wait for waterfox :) hooha alex

have you try pymake use the python side of mozbuild ???

Аватар пользователя Alex K.

Цитата:

Melanie Blower (Intel) wrote:

Your screenshot contains a message that the filesystem is out of space. This could also cause debug file corruption.

The Intel compiler recently added support for a system environment variable that pre- or post- pends compiler flags onto each compilation line. It could be useful in your case where there's a complicated build system and you're not sure how to modify all the /Zi flags into /Z7

You could set the environment variable __INTEL_POST_CFLAGS to be -Z7. Since Z7 occurs later on the command line, it will override any Zi that may occur earlier in the command line.

The filesystem has plenty of space (over 70% of 1TB free), so it can't be that. As for using /Z7 will it make much of a difference? I've never used setenv before, is it something I need to set in VS2012? Should it be added to the makefile or configure file?

Цитата:

Sergey Kostrov wrote:

>>...Hopefully it's not too annoying me posting these threads here, I've been trying for a few months unsuccessfully to
>>compile Firefox with ICC x64 and feel like I've made quite a lot of progress but I'm not quite there yet...

I assume that you're doing this just for yourself, right? Or, Is it a Firefox Community related task?

It's a bit of both.

Аватар пользователя Tim Prince

Environment variables in Windows cmd shell are set by e.g.

set __INTEL_POST_CFLAGS=Z7

or you could add "__INTEL_POST_CFLAGS=Z7" to the make command.

setenv command is available for the same purpose if you run under tcsh shell.

Аватар пользователя Alex K.

[quote=TimP (Intel)]

Environment variables in Windows cmd shell are set by e.g.

set __INTEL_POST_CFLAGS=Z7

or you could add "__INTEL_POST_CFLAGS=Z7" to the make command.

setenv command is available for the same purpose if you run under tcsh shell.

[/quote

Great, that seems to have solved that issue! But now quite a few various files that compile fine under VC and GCC, get a 'incomplete type is not allowed' error. For now I'm having VC compile them, I just hope it's not a lot of files that need to be compiled this way! Once I get a successful build I'll post a proper report.

Аватар пользователя Sergey Kostrov

>>>>...I assume that you're doing this just for yourself, right? Or, Is it a Firefox Community related task?
>>
>>It's a bit of both.

One more question: How big a complete set of sources ( in MB ) before compilation?

Аватар пользователя Alex K.

Цитата:

Sergey Kostrov wrote:

>>>>...I assume that you're doing this just for yourself, right? Or, Is it a Firefox Community related task?
>>
>>It's a bit of both.

One more question: How big a complete set of sources ( in MB ) before compilation?

Compressed, it's around 100MB. I think there's around 100,000 files, but not all get compiled. So there's bound to be compatibility issues. Here's the source tree if you're interested:
http://mxr.mozilla.org/mozilla-release/source/

Зарегистрируйтесь, чтобы оставить комментарий.