lirobot / ulib Goto Github PK

What steps will reproduce the problem?
1. tar xvzf ulib-2.0.0beta_src.tar.gz
2. cd ulib-svn
3. make

What is the expected output? What do you see instead?
lch@hyperstream:/DT/local/ulib-svn$ make
make -C src/
make[1]: Entering directory `/DT/local/ulib-svn/src'
make -C base/
make[2]: Entering directory `/DT/local/ulib-svn/src/base'
  AR    ../../lib/libbase.a
make[2]: Leaving directory `/DT/local/ulib-svn/src/base'
make -C ext1/
make[2]: Entering directory `/DT/local/ulib-svn/src/ext1'
make -C c++/
make[3]: Entering directory `/DT/local/ulib-svn/src/ext1/c++'
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext1/c++'
make -C bloom_filter/
make[3]: Entering directory `/DT/local/ulib-svn/src/ext1/bloom_filter'
  AR    ../../../lib/libbfilter.a
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext1/bloom_filter'
make -C comb/
make[3]: Entering directory `/DT/local/ulib-svn/src/ext1/comb'
  AR    ../../../lib/libcomb.a
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext1/comb'
make -C console/
make[3]: Entering directory `/DT/local/ulib-svn/src/ext1/console'
  AR    ../../../lib/libconsole.a
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext1/console'
make -C rng/
make[3]: Entering directory `/DT/local/ulib-svn/src/ext1/rng'
  AR    ../../../lib/librng.a
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext1/rng'
make[2]: Leaving directory `/DT/local/ulib-svn/src/ext1'
make -C ext2/
make[2]: Entering directory `/DT/local/ulib-svn/src/ext2'
make -C thread
make[3]: Entering directory `/DT/local/ulib-svn/src/ext2/thread'
  AR    ../../../lib/libthread.a
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext2/thread'
make -C osdep
make[3]: Entering directory `/DT/local/ulib-svn/src/ext2/osdep'
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext2/osdep'
make -C reentrant
make[3]: Entering directory `/DT/local/ulib-svn/src/ext2/reentrant'
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext2/reentrant'
make -C mapreduce
make[3]: Entering directory `/DT/local/ulib-svn/src/ext2/mapreduce'
  AR    ../../../lib/libmr.a
make[3]: Leaving directory `/DT/local/ulib-svn/src/ext2/mapreduce'
make[2]: Leaving directory `/DT/local/ulib-svn/src/ext2'
make[1]: Leaving directory `/DT/local/ulib-svn/src'
make -C lib/
make[1]: Entering directory `/DT/local/ulib-svn/lib'
make[1]: `libulib.a' is up to date.
make[1]: Leaving directory `/DT/local/ulib-svn/lib'
make -C test/
make[1]: Entering directory `/DT/local/ulib-svn/test'
-e   GEN    hash_chain_r1.test
In file included from ../include/ulib/os_typelock.h:30:0,
                 from ../include/ulib/os_regionlock.h:32,
                 from ../include/ulib/hash_chain_r.h:32,
                 from hash_chain_r1.cpp:3:
../include/ulib/os_spinlock.h: In function ‘void 
spin_lock_ticket(ticket_lock_t*)’:
../include/ulib/os_spinlock.h:302:37: error: expected primary-expression before 
‘.’ token
../include/ulib/os_spinlock.h: In function ‘void 
spin_wrlock_ticket(ticket_rwlock_t*)’:
../include/ulib/os_spinlock.h:357:39: error: expected primary-expression before 
‘.’ token
../include/ulib/os_spinlock.h: In function ‘void 
spin_wrunlock_ticket(ticket_rwlock_t*)’:
../include/ulib/os_spinlock.h:375:39: error: expected primary-expression before 
‘.’ token
../include/ulib/os_spinlock.h: In function ‘void 
spin_rdlock_ticket(ticket_rwlock_t*)’:
../include/ulib/os_spinlock.h:403:39: error: expected primary-expression before 
‘.’ token
make[1]: *** [hash_chain_r1.test] Error 1
make[1]: Leaving directory `/DT/local/ulib-svn/test'
make: *** [all] Error 2


What version of the product are you using? On what operating system?
Ubuntu 12.04.2 LTS

Please provide any additional information below.

Here the generated code. There is a ".head_tail" instead of "inc.head_tail":

static __always_inline void spin_lock_ticket(ticket_lock_t *lock)
{
    register union ticket_lock inc = { .head_tail = 1 << TICKET_SHIFT };

#ifdef LARGE_CPUSET
    inc.head_tail = atomic_fetchadd32(&lock->tickets, inc.head_tail);
#else
    inc.head_tail = atomic_fetchadd16(&lock->tickets, inc.head_tail);
#endif
[...]

What steps will reproduce the problem?
executing the test case
What is the expected output? What do you see instead?
/tmp/ccb66wEl.o: In function `ulib::thread::stop_and_join()':
/home/dgu/ExtProject/ulib-read-only/test/../include/ulib/thread.h:68: undefined 
reference to `pthread_join'


What version of the product are you using? On what operating system?

the latest
Please provide any additional information below.

What steps will reproduce the problem?
1. make
2. make test


What is the expected output? What do you see instead?
hash_align.test: hash_align.cpp:42: int main(): Assertion `months["april"] == 
30' failed.
Aborted (core dumped)
hash_multihash_r.test: hash_multihash_r.cpp:70: int main(): Assertion `map[num] 
== -1' failed.
Aborted (core dumped)
[main:6]        [W] warning log
[main:8]        [E] fatal log
median.test: median.cpp:32: int main(): Assertion `m == data[k]' failed.
Aborted (core dumped)
55 success
58 in all

What version of the product are you using? On what operating system?
$ uname -a
Linux server1 3.8.0-19-generic #29-Ubuntu SMP Wed Apr 17 18:16:28 UTC 2013 
x86_64 x86_64 x86_64 GNU/Linux

Please provide any additional information below.

What steps will reproduce the problem?
1. svn checkout http://ulib.googlecode.com/svn/trunk/ ulib
2. cd ulib
3. make
[...]
cc1: warnings being treated as errors
argv_split.c: In function ‘argv_split’:
argv_split.c:95: warning: implicit declaration of function ‘strndup’
argv_split.c:95: warning: incompatible implicit declaration of built-in 
function ‘strndup’
make[3]: *** [argv_split.o] Error 1
make[2]: *** [all] Error 2
make[1]: *** [all] Error 2
make: *** [ulib_dist.tar.gz] Error 2

What is the expected output? What do you see instead?
strndup function is missing on OSX and -L can't be used with extra space.

What version of the product are you using? On what operating system?
ulib from subversion (trunk)

Please provide any additional information below.
The following patch (ulib-osx.patch) fix all the compile problem on OSX 
(10.6.8+).

What steps will reproduce the problem?
1. Use aligned has of std::string to int or a custom string class to int; 
compilers Win/Vis Studio 11 or linux/gcc 4.4.6 (the std gcc on latest redhat).
2. I get a crash right away: on first cal of insert. 
3. Debugging is near impossible, due to most code being in #define macros! 

What is the expected output? What do you see instead?


What version of the product are you using? On what operating system?


Please provide any additional information below.

Silly question: I was just interested to know if this C++ class is tested and 
works correctly without mem corruption or leaks? Or do you only use/test the C 
version? Looks e.g. that code assignes into area that is only allocated by C, 
so are we sure construction/destruction for objects works? 

Purpose of code changes on this branch:
add gtest.h

When reviewing my code changes, please focus on:


After the review, I'll merge this branch into:
/trunk

What steps will reproduce the problem?
1. define an open_hash_set<uint64_t> as my_set, then insert [0,1<<26) as keys 
into it by step N in a loop
2. define an open_hash_map<uint64_t, uint64_t> as my_map
3. iterate the open_hash_set and get iterator's key() as my_map's key, then set 
my_map[key] = 1
4. change loop's step N, and watch the costing time of my_map's operation 

What is the expected output? What do you see instead?
Expected output: costing time should be less and less as N increasing
I see: the costing time is not correct(more than 20 seconds when N is 3)  

What version of the product are you using? On what operating system?
Version: 2.0.1
OS: debian7 64bit, gcc 4.7.2 

Please provide any additional information below.

What steps will reproduce the problem?

1. make

What is the expected output? What do you see instead?

ld: library not found for -lrt
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[1]: *** [aes.test] Error 1
make: *** [all] Error 2

What version of the product are you using? On what operating system?
2.0.1

Please provide any additional information below.
I am using OSX Mavericks with latest Xcode command line tools

Hi 

i just downloaded the source tar from site 
wget http://ulib.googlecode.com/files/ulib-2.0.1_src.tar.gz

i checked the checksum its correct.
i extracted the file and did make and i found following issue 
i am using Ubuntu 13.04 over Intel® Core™ i7-2630QM CPU @ 2.00GHz × 8 

below is the problem listed.


amey@axps:~/work/installables/ulib-svn$ make
make -C src/
make[1]: Entering directory `/home/amey/work/installables/ulib-svn/src'
make -C base/
make[2]: Entering directory `/home/amey/work/installables/ulib-svn/src/base'
-e   CC bitmap.c
-e   CC crypt_aes.c
-e   CC crypt_md5.c
-e   CC crypt_rc4.c
-e   CC crypt_sha1.c
-e   CC crypt_sha256.c
-e   CC hash_func.c
hash_func.c: In function ‘hash_ferm64’:
hash_func.c:92:3: error: impossible register constraint in ‘asm’
hash_func.c:114:1: error: impossible register constraint in ‘asm’
hash_func.c:106:3: error: impossible register constraint in ‘asm’
hash_func.c:114:1: error: impossible register constraint in ‘asm’
make[2]: *** [hash_func.o] Error 1
make[2]: Leaving directory `/home/amey/work/installables/ulib-svn/src/base'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/amey/work/installables/ulib-svn/src'
make: *** [all] Error 2

What steps will reproduce the problem?
1. Just use a 2-block 256-bit AES encrypted input; Vis Studio 10/11, 64-bit. 
Sandybridge PC.

2. Call code like below (iv is 16 bytes):
    AES_KEY aeskey;
    AES_set_decrypt_key( key, 32 * 8, &aeskey );
    AES_cbc_decrypt( ( const unsigned char * ) in, ( unsigned char * ) in, iv, 2, &aeskey );

"in" is allocated with malloc, which guarantees 16-byte alignment on 64-bit.
"iv" is:
uchar iv[] = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 };


What is the expected output? What do you see instead?
This decrypts all bytes in the 1st block correctly. 
But the 2nd block gets garbage.
When i use a different buffer (non-overlapping) for output, it works fine.


What version of the product are you using? On what operating system?
Latest


Please provide any additional information below.
I also tried to add a couple  of changes, but neither of them helped:


(1)

Looking at original code you may need this  too:

#ifdef _MSC_VER
#define SWAP(x) (_lrotl(x, 8) & 0x00ff00ff | _lrotr(x, 8) & 0xff00ff00)
#define GETU32(p) SWAP(*((u32 *)(p)))
#define PUTU32(ct, st) { *((u32 *)(ct)) = SWAP((st)); }
#else

#define GETU32(pt) (((u32)(pt)[0] << 24) ^ ((u32)(pt)[1] << 16) ^ ((u32)(pt)[2] 
<<  8) ^ ((u32)(pt)[3]))
#define PUTU32(ct, st) { (ct)[0] = (u8)((st) >> 24); (ct)[1] = (u8)((st) >> 
16); (ct)[2] = (u8)((st) >>  8); (ct)[3] = (u8)(st); }
#endif


(2)
YTou do also need this change:

#if defined(__x86_64__) || defined(_M_X64) 

#define AES_BLOCK_OP(dst, src, op) AES_BLOCK_OP64(dst, src, op)
#else
#define AES_BLOCK_OP(dst, src, op) AES_BLOCK_OP32(dst, src, op)
#endif


FYI, i also tried code built in 32-bit - same issue.

Note: VS Express version is available freely and you could also get Beta for VS 
11.



Possible to fix? As the called function says it works on overlapping buffers, i 
assume the cbc can do the same?