# Nginx-0.8.53 segfault

**URL:** <https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687>\
**Category:** NGINX\
**Created:** [January 12, 2011, 10:24pm UTC](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687 "2011-01-12T22:24:54Z")\
**Posts on this page:** 5\
**Page:** 2

<div class="post-metadata">

**Author:** ![maxim\_c](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/3bc359/32.png) [@maxim\_c](https://www.ruby-forum.com/u/maxim_c)\
**Post date:** [January 14, 2011, 8:02pm UTC](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687/21 "2011-01-14T20:02:15Z")

</div>

On 1/14/11 1:55 PM, “Maxim D.” [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote:

> Right now I think correct solution would be to move r-\>cache into  
> upstream’s private data (r-\>upstream). Igor?

FWIW, we are doing some thing with file\_cache that are not upstream  
based.  
I hope to be able to release those soon.

–  
Brian A.

---

<div class="post-metadata">

**Author:** ![maxim\_c](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/3bc359/32.png) [@maxim\_c](https://www.ruby-forum.com/u/maxim_c)\
**Post date:** [January 14, 2011, 9:06pm UTC](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687/22 "2011-01-14T21:06:28Z")

</div>

Thank you guys!  
Is there some workaround? Should I disable caching of images produced by  
image filter?

---

<div class="post-metadata">

**Author:** ![maxim\_c](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/3bc359/32.png) [@maxim\_c](https://www.ruby-forum.com/u/maxim_c)\
**Post date:** [January 14, 2011, 11:25pm UTC](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687/23 "2011-01-14T23:25:58Z")

</div>

Maxim, thank you again!

It worked for me.

---

<div class="post-metadata">

**Author:** ![maxim\_c](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/3bc359/32.png) [@maxim\_c](https://www.ruby-forum.com/u/maxim_c)\
**Post date:** [January 14, 2011, 10:00pm UTC](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687/24 "2011-01-14T22:00:28Z")

</div>

Hello!

On Fri, Jan 14, 2011 at 09:06:28PM +0100, Maxim C. wrote:

> Thank you guys!  
> Is there some workaround?

Using named location in error\_page 415 should help, i.e.

```
location ... {
    ...
    image_filter ...
    error_page 415 = @empty;
}

location @empty {
    empty_gif;
}

```

Maxim D.

---

<div class="post-metadata">

**Author:** ![maxim\_c](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/m/3bc359/32.png) [@maxim\_c](https://www.ruby-forum.com/u/maxim_c)\
**Post date:** [January 20, 2011, 12:25am UTC](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687/25 "2011-01-20T00:25:50Z")

</div>

Hello!

On Fri, Jan 14, 2011 at 09:55:35PM +0300, Maxim D. wrote:

> > 790 if (c-\>updated) {
> 
> ```
> location / {
> }
> 
> location /bad/ {
> return 404;
> }
> }
> 
> ```
> 
> Right now I think correct solution would be to move r-\>cache into  
> upstream’s private data (r-\>upstream). Igor?

The attached patch moves r-\>cache into r-\>upstream (and resolves  
the SIGSEGV in question).

To Igor:

The patch is rather big and touches lots of code. I would like to  
get some explicit response on it to avoid unneeded merge work.

Maxim D.

[Previous page](https://www.ruby-forum.com/t/nginx-0-8-53-segfault/201687.md?page=1)