apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rainer Jung <rainer.j...@kippdata.de>
Subject Re: Appropriate patches for 2.2.19 and 2.0.64?
Date Sat, 10 Sep 2011 17:21:39 GMT
On 10.09.2011 15:00, William A. Rowe Jr. wrote:
> On 9/6/2011 4:24 PM, William A. Rowe Jr. wrote:
>> On 9/6/2011 9:01 AM, Jeff Trawick wrote:
>>>
>>> I can take care of T&R of APR 0.9.20 in a couple of days provided that
>>> someone first gets CVE-2011-1928 resolved in that branch.
>>
>> That's already done for some time now;
>>
>>   http://svn.apache.org/viewvc/apr/apr/branches/0.9.x/strings/
>>
>> although I still need to update CHANGES already.
> 
> Now changed, and rechecked our merged rev.  This looks good to go
> if you would like to move forwards.
> 
> Did anyone else have immediate concerns for this 0.9 branch?

I just did a quick check for the current 0.9.x heads:

1) APR
======

make check (testall) crashes when buildingout of tree, because test/data
isn't found. If I copy it in, I get one failure:

Failed tests in Socket Creation:
1) sendto_receivefrom: expected <0> but was <126>

But that's not a regression from 0.9.19.

2) APU-UTIL
===========

Builds and checks fine.


3) APR-ICONV
============

Build fails:

.../r1167521/./lib/iconv_ces_euc.c: In function 'apr_iconv_euc_open':
.../r1167521/./lib/iconv_ces_euc.c:65: error: invalid lvalue in assignment

Line is:

CESTOSTATE(ces) = state;

which is

((iconv_ces_euc_state_t *)(ces)->data) = state;

It is old code, so probably not a regression. IMHO the fix would be:

Index:
/shared/build/dev/httpd/sources/apr-iconv/0.9.x/r1167521/lib/iconv_ces_euc.c
===================================================================
---
/shared/build/dev/httpd/sources/apr-iconv/0.9.x/r1167521/lib/iconv_ces_euc.c
       (revision 1167521)
+++
/shared/build/dev/httpd/sources/apr-iconv/0.9.x/r1167521/lib/iconv_ces_euc.c
       (working copy)
@@ -62,7 +62,7 @@
        state->nccs = ces->mod->im_depcnt;
        for (i = ces->mod->im_depcnt; i; i--, depmod = depmod->im_next)
                state->ccs[i - 1] = depmod;
-       CESTOSTATE(ces) = state;
+       ces->data = (void *)state;
        return APR_SUCCESS;
 }

and then again

.../r1167521/./ces/ucs2-internal.c: In function 'convert_from_ucs':
.../r1167521/./ces/ucs2-internal.c:62: error: invalid lvalue in increment
.../r1167521/./ces/ucs2-internal.c: In function 'convert_to_ucs':
.../r1167521/./ces/ucs2-internal.c:74: error: invalid lvalue in increment

Haven't checked those for fixes yet. Need to rush for something to eat
now :)

Regards,

Rainer

Mime
View raw message