Account Options

  1. Sign in
The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Google Groups Home
« Groups Home
google-perftools
Home
+ new post
About this group
Join this group
Message from discussion Mipsel platform support
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
Gayathri Vijayakumar  
View profile  
 More options Jul 19 2012, 7:39 am
From: Gayathri Vijayakumar <gayathrivijayaku...@gmail.com>
Date: Thu, 19 Jul 2012 04:39:54 -0700 (PDT)
Local: Thurs, Jul 19 2012 7:39 am
Subject: Re: Mipsel platform support
Print | View thread | Show original | Report this message | Find messages by this author

I have been trying to compile gperftools version 1.8.2 (as I could find
MIPS patch available only for this version) for mipsel-linux-uclibc. I had
to make certain chnages here and there and I am able to compile. I am able
to use CPUprofiler by linking libprofiler. But the tcmalloc fails me
miserably. Though I could compile it, when linking it to my test app and
running it, I see that it seg faults. Analysing the core dump, I see that
the Delete funstion fails. I am pasting the coredump here. Any help would
be great! Thanks in advance folks.

#0  tcmalloc::PageHeapAllocator<tcmalloc::StackTrace>::Delete () at
src/page_heap_allocator.h:88
#1  tcmalloc::PageHeapAllocator<tcmalloc::StackTrace>::Init () at
src/page_heap_allocator.h:58
#2  tcmalloc::Static::InitStaticVars () at src/static_vars.cc:56
#3  0x2ac6c234 in tcmalloc::ThreadCache::InitModule () at
src/thread_cache.cc:311
#4  0x2ac58178 in tcmalloc::ThreadCache::GetCache (size=320) at
src/thread_cache.h:380
#5  do_malloc (size=320) at src/tcmalloc.cc:999
#6  do_malloc_or_cpp_alloc (size=320) at src/tcmalloc.cc:946
#7  0x2ac82780 in tc_realloc (old_ptr=0x0, new_size=320) at
src/tcmalloc.cc:1450
#8  0x2ad6bbf8 in ?? () from /lib/libc.so.0

(gdb) fr 0
#0  tcmalloc::PageHeapAllocator<tcmalloc::StackTrace>::Delete () at
src/page_heap_allocator.h:88
88          *(reinterpret_cast<void**>(p)) = free_list_;
(gdb) p free_list_
Cannot access memory at address 0x8

My build commands were:

./configure --host=mipsel-linuc-uclibc
make
make install

Thanks,
Gayathri

On Friday, May 18, 2012 5:54:46 AM UTC+5:30, Darren Garvey wrote:

> Hi,

> I've made a bit of an attempt recently to compile tcmalloc for a mipsel
> platform and not had any immediate success. I've seen a few mentions by
> others that they've had tcmalloc working on mipsel (albeit with no public
> patches), along with a few mentions that mipsel support is pending in
> mainline, but /not yet/.

> Before going headstrong into it and attempting to redo the work of others,
> I wanted to know if I'm missing anything obvious as far as mipsel support
> goes? Do others have this working? Would they be willing to share their
> patches, or have they already with some public git repo, for instance?

> I'm happy putting effort into getting this working, but hate the idea of
> reinventing work that others have already done.

> Cheers,

> Darren

On Friday, May 18, 2012 5:54:46 AM UTC+5:30, Darren Garvey wrote:

> Hi,

> I've made a bit of an attempt recently to compile tcmalloc for a mipsel
> platform and not had any immediate success. I've seen a few mentions by
> others that they've had tcmalloc working on mipsel (albeit with no public
> patches), along with a few mentions that mipsel support is pending in
> mainline, but /not yet/.

> Before going headstrong into it and attempting to redo the work of others,
> I wanted to know if I'm missing anything obvious as far as mipsel support
> goes? Do others have this working? Would they be willing to share their
> patches, or have they already with some public git repo, for instance?

> I'm happy putting effort into getting this working, but hate the idea of
> reinventing work that others have already done.

> Cheers,

> Darren


 
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
©2013 Google