map returns all undefs for PERL_HASH_FUNC_DJB2 miniperl #13347
Comments
From @bulk88Created by @bulk88Using PERL_HASH_FUNC_DJB2 with miniperl, map() passed all undef $_s to sample script my @arr = ('num1','num2','num3'); with DJB2 miniperl Use of uninitialized value $_ in concatenation (.) or string at Use of uninitialized value $_ in concatenation (.) or string at C:\Documents and Settings\Owner\Desktop\cpan libs\p519\perl\win32> with regular/default hash perl 5.10 ___________________________________________________ C:\Documents and Settings\Owner\Desktop\cpan libs\p519\perl\win32> Patch of hv_func.h that shows how I switched hash funcs is attached. This test script was cut down from write_buildcustomize.pl failing Perl Info |
From @bulk88hv_func change.patchdiff --git "a/C:\\WINDOWS\\TEMP\\hv_69.tmp\\hv_func-b68084b-left.h" "b/C:\\Documents and Settings\\Owner\\Desktop\\cpan libs\\p519\\perl\\hv_func.h"
index 191912a..28b8367 100644
--- "a/C:\\WINDOWS\\TEMP\\hv_69.tmp\\hv_func-b68084b-left.h"
+++ "b/C:\\Documents and Settings\\Owner\\Desktop\\cpan libs\\p519\\perl\\hv_func.h"
@@ -22,7 +22,8 @@
|| defined(PERL_HASH_FUNC_ONE_AT_A_TIME_HARD) \
|| defined(PERL_HASH_FUNC_ONE_AT_A_TIME_OLD) \
)
-#define PERL_HASH_FUNC_ONE_AT_A_TIME_HARD
+//#define PERL_HASH_FUNC_ONE_AT_A_TIME_HARD
+#define PERL_HASH_FUNC_DJB2
#endif
#if defined(PERL_HASH_FUNC_SIPHASH)
|
From @bulk88C:\Documents and Settings\Owner\Desktop\cpan libs\p519\perl\win32>..\miniperl.ex C:\Documents and Settings\Owner\Desktop\cpan libs\p519\perl\win32> |
From @bulk88Changing the define to "#define PERL_HASH_FUNC_ONE_AT_A_TIME_OLD" does -- |
From @bulk88[16:32] <@rurban> PERL_HASH_FUNC_SDBM and PERL_HASH_FUNC_DJB2: simple fix: U32 hash = *((U32*)seed); untested by me, put here for safe keeping -- |
From zefram@fysh.orgbulk88 via RT wrote:
This apparently refers to the two lines U32 hash = *((U32*)seed + len); in hv_func.h, in S_perl_hash_sdbm() and S_perl_hash_djb2(), which look U32 hash = *((U32*)seed) + len; as seen in S_perl_hash_superfast() and two others. -zefram |
|
The RT System itself - Status changed from 'new' to 'open' |
From @rurbanOn Sat Apr 12 03:43:15 2014, zefram@fysh.org wrote:
sorry, but this is nonsense.
-- |
From zefram@fysh.orgReini Urban via RT wrote:
Why is it wrong to add the length into the hash at that point? -zefram |
From @rurbanOn Sat Apr 12 10:50:44 2014, zefram@fysh.org wrote:
We use a global hash seed. It makes no sense to mix that seed with the This is my current version of the patch |
From zefram@fysh.orgReini Urban via RT wrote:
I don't see how the scope of the seed is relevant to how the key length In the case of these two hash functions, DJB2 and SDBM, including the I don't know the real reason why these functions include the length. -zefram |
From @demerphqOn 12 April 2014 22:02, Zefram <zefram@fysh.org> wrote:
Your analysis is correct. It was made to harden these functions Yves -- |
From @demerphqOn 12 April 2014 19:47, Reini Urban via RT <perlbug-followup@perl.org> wrote:
No it is not. It is the correct fix to my change.
Adding the length is not 100% faithful to the original algorithm. See Zeframs mails for details.
No, don't. I will push patch for this, and it will look like Zefram posted.
I have no idea what you are talking about. Yves -- |
From @rurbanOn Apr 13, 2014, at 5:12 AM, demerphq <demerphq@gmail.com> wrote:
It is nonsense nevertheless, but it is the correct fix for the seed=0 case. The real fix for collision attacks is to avoid O(n/2) collision lookup.
Bad, because you wrote it and blogged about. To refresh your mind: |
From @tonycozOn Sat Apr 12 03:43:15 2014, zefram@fysh.org wrote:
This was fixed by Yves's 54e07e2. Closing. Tony |
|
@tonycoz - Status changed from 'open' to 'resolved' |
Migrated from rt.perl.org#120208 (status was 'resolved')
Searchable as RT120208$
The text was updated successfully, but these errors were encountered: