FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER … #320

laurynas-biveinis · 2015-09-10T07:58:30Z

…is wrong for debug builds)

Revert the bits of FT-699 that cause a debug build crash in
toku_mutex_init because of passing an initialised mutex where an
uninitialised one is expected.

…is wrong for debug builds) Revert the bits of FT-699 that cause a debug build crash in toku_mutex_init because of passing an initialised mutex where an uninitialised one is expected.

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER …

kuszmaul · 2015-09-10T17:15:43Z

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis <notifications@github.com

wrote:

Merged #320 #320.

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

george-lorch · 2015-09-10T17:22:14Z

It's broken and being fixed, tested and verified right now.

On 09/10/2015 10:15 AM, Bradley C. Kuszmaul wrote:

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis
<notifications@github.com

wrote:

Merged #320 #320.

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

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

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

kuszmaul · 2015-09-10T17:35:40Z

OK. Incidentally I made a pull request (it's on tokutek/ft-index... does
that get mirrored back to percona? Should I stop using tokutek/ft-index?)

In my pull request, I not only switched it back, but I also wrote a bunch
of explanation about what's going on.

-Bradley

On Thu, Sep 10, 2015 at 1:22 PM, georgelorchpercona <
notifications@github.com> wrote:

It's broken and being fixed, tested and verified right now.

On 09/10/2015 10:15 AM, Bradley C. Kuszmaul wrote:

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis
<notifications@github.com

wrote:

Merged #320 #320.

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

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

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

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

george-lorch · 2015-09-10T17:49:14Z

No, it does not get mirrored back unfortunately. ft-index in the Tokutek
repo needs to be closed off.

I have a similar fix you can use the use the PTHREAD_MUTEX_INITIALIZER
macro if C++11, but still not particularly good form.

On 09/10/2015 10:35 AM, Bradley C. Kuszmaul wrote:

OK. Incidentally I made a pull request (it's on tokutek/ft-index... does
that get mirrored back to percona? Should I stop using tokutek/ft-index?)

In my pull request, I not only switched it back, but I also wrote a bunch
of explanation about what's going on.

-Bradley

On Thu, Sep 10, 2015 at 1:22 PM, georgelorchpercona <
notifications@github.com> wrote:

It's broken and being fixed, tested and verified right now.

On 09/10/2015 10:15 AM, Bradley C. Kuszmaul wrote:

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis
<notifications@github.com

wrote:

Merged #320 #320.

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

—
Reply to this email directly or view it on GitHub

#320 (comment).

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

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

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

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

kuszmaul · 2015-09-10T18:00:59Z

Is there a branch with the candidate fix?

Bradley C Kuszmaul - via snartphone
On Sep 10, 2015 1:49 PM, "georgelorchpercona" notifications@github.com
wrote:

No, it does not get mirrored back unfortunately. ft-index in the Tokutek
repo needs to be closed off.

I have a similar fix you can use the use the PTHREAD_MUTEX_INITIALIZER
macro if C++11, but still not particularly good form.

On 09/10/2015 10:35 AM, Bradley C. Kuszmaul wrote:

OK. Incidentally I made a pull request (it's on tokutek/ft-index... does
that get mirrored back to percona? Should I stop using tokutek/ft-index?)

In my pull request, I not only switched it back, but I also wrote a bunch
of explanation about what's going on.

-Bradley

On Thu, Sep 10, 2015 at 1:22 PM, georgelorchpercona <
notifications@github.com> wrote:

It's broken and being fixed, tested and verified right now.

On 09/10/2015 10:15 AM, Bradley C. Kuszmaul wrote:

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis
<notifications@github.com

wrote:

Merged #320 #320.

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

—
Reply to this email directly or view it on GitHub

#320 (comment).

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

—
Reply to this email directly or view it on GitHub
<#320 (comment)
.

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

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

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

george-lorch · 2015-09-10T18:03:44Z

Yeah, I took your fix instead of mine which was similar as yours had
better commenting. I'll have the master updated shortly once I've
finished testing, maybe another 15 minutes or so.

On 09/10/2015 11:01 AM, Bradley C. Kuszmaul wrote:

Is there a branch with the candidate fix?

Bradley C Kuszmaul - via snartphone
On Sep 10, 2015 1:49 PM, "georgelorchpercona" notifications@github.com
wrote:

No, it does not get mirrored back unfortunately. ft-index in the Tokutek
repo needs to be closed off.

I have a similar fix you can use the use the PTHREAD_MUTEX_INITIALIZER
macro if C++11, but still not particularly good form.

On 09/10/2015 10:35 AM, Bradley C. Kuszmaul wrote:

OK. Incidentally I made a pull request (it's on
tokutek/ft-index... does
that get mirrored back to percona? Should I stop using
tokutek/ft-index?)

In my pull request, I not only switched it back, but I also wrote
a bunch
of explanation about what's going on.

-Bradley

On Thu, Sep 10, 2015 at 1:22 PM, georgelorchpercona <
notifications@github.com> wrote:

It's broken and being fixed, tested and verified right now.

On 09/10/2015 10:15 AM, Bradley C. Kuszmaul wrote:

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead
of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis
<notifications@github.com

wrote:

Merged #320 #320.

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

—
Reply to this email directly or view it on GitHub

#320 (comment).

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

—
Reply to this email directly or view it on GitHub

<#320 (comment)
.

—
Reply to this email directly or view it on GitHub

#320 (comment).

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

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

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

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

laurynas-biveinis · 2015-09-10T18:08:30Z

Sorry about that and thanks for the catch. The commit message of FT-700 is what I tried to do, but apparently messed up.

george-lorch · 2015-09-10T18:21:50Z

OK Bradley, master is up to date, sorry about the mess.

On 09/10/2015 11:01 AM, Bradley C. Kuszmaul wrote:

Is there a branch with the candidate fix?

Bradley C Kuszmaul - via snartphone
On Sep 10, 2015 1:49 PM, "georgelorchpercona" notifications@github.com
wrote:

No, it does not get mirrored back unfortunately. ft-index in the Tokutek
repo needs to be closed off.

I have a similar fix you can use the use the PTHREAD_MUTEX_INITIALIZER
macro if C++11, but still not particularly good form.

On 09/10/2015 10:35 AM, Bradley C. Kuszmaul wrote:

OK. Incidentally I made a pull request (it's on
tokutek/ft-index... does
that get mirrored back to percona? Should I stop using
tokutek/ft-index?)

In my pull request, I not only switched it back, but I also wrote
a bunch
of explanation about what's going on.

-Bradley

On Thu, Sep 10, 2015 at 1:22 PM, georgelorchpercona <
notifications@github.com> wrote:

It's broken and being fixed, tested and verified right now.

On 09/10/2015 10:15 AM, Bradley C. Kuszmaul wrote:

master is still broken (it uses TOKU_MUTEX_INITIALIZER instead
of of
ZERO_MUTEX_INITIALIZER). What's the story with this patch?

On Thu, Sep 10, 2015 at 3:59 AM, Laurynas Biveinis
<notifications@github.com

wrote:

Merged #320 #320.

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

—
Reply to this email directly or view it on GitHub

#320 (comment).

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

—
Reply to this email directly or view it on GitHub

<#320 (comment)
.

—
Reply to this email directly or view it on GitHub

#320 (comment).

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

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

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

George O. Lorch III
Software Engineer, Percona
US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

laurynas-biveinis · 2015-09-14T14:45:35Z

ZERO_COND_INITIALIZER needs the same treatment, logged FT-702, will submit a PR shortly

kuszmaul · 2015-09-14T16:05:31Z

I believe that d8598d5 fixes that, but
it's on my side branch.

-Bradley

On Mon, Sep 14, 2015 at 10:45 AM, Laurynas Biveinis <
notifications@github.com> wrote:

ZERO_COND_INITIALIZER needs the same treatment, logged FT-702, will submit
a PR shortly

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

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER …

00f4151

…is wrong for debug builds) Revert the bits of FT-699 that cause a debug build crash in toku_mutex_init because of passing an initialised mutex where an uninitialised one is expected.

laurynas-biveinis added a commit that referenced this pull request Sep 10, 2015

Merge pull request #320 from laurynas-biveinis/ft-700

596efd7

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER …

laurynas-biveinis merged commit 596efd7 into percona:master Sep 10, 2015

laurynas-biveinis deleted the ft-700 branch September 10, 2015 07:59

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER … #320

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER … #320

laurynas-biveinis commented Sep 10, 2015

kuszmaul commented Sep 10, 2015

george-lorch commented Sep 10, 2015

kuszmaul commented Sep 10, 2015

george-lorch commented Sep 10, 2015

kuszmaul commented Sep 10, 2015

george-lorch commented Sep 10, 2015

laurynas-biveinis commented Sep 10, 2015

george-lorch commented Sep 10, 2015

laurynas-biveinis commented Sep 14, 2015

kuszmaul commented Sep 14, 2015

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER … #320

FT-700 (Replacing ZERO_MUTEX_INITIALIZER with TOKU_MUTEX_INITIALIZER … #320

Conversation

laurynas-biveinis commented Sep 10, 2015

kuszmaul commented Sep 10, 2015

george-lorch commented Sep 10, 2015

kuszmaul commented Sep 10, 2015

george-lorch commented Sep 10, 2015

kuszmaul commented Sep 10, 2015

george-lorch commented Sep 10, 2015

laurynas-biveinis commented Sep 10, 2015

george-lorch commented Sep 10, 2015

laurynas-biveinis commented Sep 14, 2015

kuszmaul commented Sep 14, 2015