summaryrefslogtreecommitdiff
path: root/block-sha1
AgeCommit message (Collapse)AuthorFilesLines
2012-07-23Merge branch 'jn/block-sha1'Libravatar Junio C Hamano1-4/+4
The code to load a word one-byte-at-a-time was optimized into a word-wide load instruction even when the pointer was not aligned, which caused issues on architectures that do not like unaligned access. * jn/block-sha1: Makefile: BLK_SHA1 does not require fast htonl() and unaligned loads block-sha1: put expanded macro parameters in parentheses block-sha1: avoid pointer conversion that violates alignment constraints
2012-07-22block-sha1: put expanded macro parameters in parenthesesLibravatar Jonathan Nieder1-2/+2
't' is currently always a numeric constant, but it can't hurt to prepare for the day that it becomes useful for a caller to pass in a more complex expression. Suggested-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2012-07-22block-sha1: avoid pointer conversion that violates alignment constraintsLibravatar Jonathan Nieder1-3/+3
With 660231aa (block-sha1: support for architectures with memory alignment restrictions, 2009-08-12), blk_SHA1_Update was modified to access 32-bit chunks of memory one byte at a time on arches that prefer that: #define get_be32(p) ( \ (*((unsigned char *)(p) + 0) << 24) | \ (*((unsigned char *)(p) + 1) << 16) | \ (*((unsigned char *)(p) + 2) << 8) | \ (*((unsigned char *)(p) + 3) << 0) ) The code previously accessed these values by just using htonl(*p). Unfortunately, Michael noticed on an Alpha machine that git was using plain 32-bit reads anyway. As soon as we convert a pointer to int *, the compiler can assume that the object pointed to is correctly aligned as an int (C99 section 6.3.2.3 "pointer conversions" paragraph 7), and gcc takes full advantage by using a single 32-bit load, resulting in a whole bunch of unaligned access traps. So we need to obey the alignment constraints even when only dealing with pointers instead of actual values. Do so by changing the type of 'data' to void *. This patch renames 'data' to 'block' at the same time to make sure all references are updated to reflect the new type. Reported-tested-and-explained-by: Michael Cree <mcree@orcon.net.nz> Signed-off-by: Jonathan Nieder <jrnieder@gmail.com> Acked-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-06-27msvc: Select the "fast" definition of the {get,put}_be32() macrosLibravatar Ramsay Jones1-0/+1
On Intel machines, the msvc compiler defines the CPU architecture macros _M_IX86 and _M_X64 (equivalent to __i386__ and __x86_64__ respectively). Use these macros in the pre-processor expression to select the "fast" definition of the {get,put}_be32() macros. Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Acked-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-06-25msvc: Fix some compiler warningsLibravatar Ramsay Jones1-4/+4
In particular, using the normal (or production) compiler warning level (-W3), msvc complains as follows: .../sha1.c(244) : warning C4018: '<' : signed/unsigned mismatch .../sha1.c(270) : warning C4244: 'function' : conversion from \ 'unsigned __int64' to 'unsigned long', possible loss of data .../sha1.c(271) : warning C4244: 'function' : conversion from \ 'unsigned __int64' to 'unsigned long', possible loss of data Note that gcc issues a similar complaint about line 244 when compiling with -Wextra. Signed-off-by: Ramsay Jones <ramsay@ramsay1.demon.co.uk> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-18remove ARM and Mozilla SHA1 implementationsLibravatar Nicolas Pitre2-4/+8
They are both slower than the new BLK_SHA1 implementation, so it is pointless to keep them around. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-18block-sha1: guard gcc extensions with __GNUC__Libravatar Nicolas Pitre1-2/+2
With this, the code should now be portable to any C compiler. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-18make sure byte swapping is optimal for gitLibravatar Nicolas Pitre1-2/+2
We rely on ntohl() and htonl() to perform byte swapping in many places. However, some platforms have libraries providing really poor implementations of those which might cause significant performance issues, especially with the block-sha1 code. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-18block-sha1: make the size member first in the context structLibravatar Nicolas Pitre1-1/+1
This is a 64-bit value, hence having it first provides a better alignment. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-14block-sha1/sha1.c: silence compiler complaints by casting void * to char *Libravatar Brandon Casey1-2/+2
Some compilers produce errors when arithmetic is attempted on pointers to void. We want computations done on byte addresses, so cast them to char * to work them around. Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-13block-sha1: more good unaligned memory access candidatesLibravatar Nicolas Pitre1-1/+4
In addition to X86, PowerPC and S390 are capable of unaligned memory accesses. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-12block-sha1: support for architectures with memory alignment restrictionsLibravatar Nicolas Pitre1-2/+30
This is needed on architectures with poor or non-existent unaligned memory support and/or no fast byte swap instruction (such as ARM) by using byte accesses to memory and shifting the result together. This also makes the code portable, therefore the byte access methods are the defaults. Any architecture that properly supports unaligned word accesses in hardware simply has to enable the alternative methods. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-12block-sha1: split the different "hacks" to be individually selectedLibravatar Nicolas Pitre1-5/+18
This is to make it easier for them to be selected individually depending on the architecture instead of the other way around i.e. having each architecture select a list of hacks up front. That makes for clearer documentation as well. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-12block-sha1: move code aroundLibravatar Nicolas Pitre1-69/+60
Move the code around so specific architecture hacks are defined first. Also make one line comments actually one line. No code change. Signed-off-by: Nicolas Pitre <nico@cam.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-10block-sha1: improve code on large-register-set machinesLibravatar Linus Torvalds1-1/+24
For x86 performance (especially in 32-bit mode) I added that hack to write the SHA1 internal temporary hash using a volatile pointer, in order to get gcc to not try to cache the array contents. Because gcc will do all the wrong things, and then spill things in insane random ways. But on architectures like PPC, where you have 32 registers, it's actually perfectly reasonable to put the whole temporary array[] into the register set, and gcc can do so. So make the 'volatile unsigned int *' cast be dependent on a SMALL_REGISTER_SET preprocessor symbol, and enable it (currently) on just x86 and x86-64. With that, the routine is fairly reasonable even when compared to the hand-scheduled PPC version. Ben Herrenschmidt reports on a G5: * Paulus asm version: about 3.67s * Yours with no change: about 5.74s * Yours without "volatile": about 3.78s so with this the C version is within about 3% of the asm one. And add a lot of commentary on what the heck is going on. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-07block-sha1: improved SHA1 hashingLibravatar Linus Torvalds1-3/+4
I think I have found a way to avoid the gcc crazyness. Lookie here: # TIME[s] SPEED[MB/s] rfc3174 5.094 119.8 rfc3174 5.098 119.7 linus 1.462 417.5 linusas 2.008 304 linusas2 1.878 325 mozilla 5.566 109.6 mozillaas 5.866 104.1 openssl 1.609 379.3 spelvin 1.675 364.5 spelvina 1.601 381.3 nettle 1.591 383.6 notice? I outperform all the hand-tuned asm on 32-bit too. By quite a margin, in fact. Now, I didn't try a P4, and it's possible that it won't do that there, but the 32-bit code generation sure looks impressive on my Nehalem box. The magic? I force the stores to the 512-bit hash bucket to be done in order. That seems to help a lot. The diff is trivial (on top of the "rename registers with cpp" patch), as appended. And it does seem to fix the P4 issues too, although I can obviously (once again) only test Prescott, and only in 64-bit mode: # TIME[s] SPEED[MB/s] rfc3174 1.662 36.73 rfc3174 1.64 37.22 linus 0.2523 241.9 linusas 0.4367 139.8 linusas2 0.4487 136 mozilla 0.9704 62.9 mozillaas 0.9399 64.94 that's some really impressive improvement. All from just saying "do the stores in the order I told you to, dammit!" to the compiler. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-07block-sha1: perform register rotation using cppLibravatar Linus Torvalds1-27/+90
Instead of letting the compiler to figure out the optimal way to rotate register usage, explicitly rotate the register names with cpp. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: get rid of redundant 'lenW' contextLibravatar Linus Torvalds2-11/+7
.. and simplify the ctx->size logic. We now count the size in bytes, which means that 'lenW' was always just the low 6 bits of the total size, so we don't carry it around separately any more. And we do the 'size in bits' shift at the end. Suggested by Nicolas Pitre and linux@horizon.com. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: Use '(B&C)+(D&(B^C))' instead of '(B&C)|(D&(B|C))' in round 3Libravatar Linus Torvalds1-1/+1
It's an equivalent expression, but the '+' gives us some freedom in instruction selection (for example, we can use 'lea' rather than 'add'), and associates with the other additions around it to give some minor scheduling freedom. Suggested-by: linux@horizon.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: macroize the rounds a bit furtherLibravatar Linus Torvalds1-30/+26
Avoid repeating the shared parts of the different rounds by adding a macro layer or two. It was already more cpp than C. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: re-use the temporary array as we calculate the SHA1Libravatar Linus Torvalds1-12/+16
The mozilla-SHA1 code did this 80-word array for the 80 iterations. But the SHA1 state is really just 512 bits, and you can actually keep it in a kind of "circular queue" of just 16 words instead. This requires us to do the xor updates as we go along (rather than as a pre-phase), but that's really what we want to do anyway. This gets me really close to the OpenSSL performance on my Nehalem. Look ma, all C code (ok, there's the rol/ror hack, but that one doesn't strictly even matter on my Nehalem, it's just a local optimization). Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: make the 'ntohl()' part of the first SHA1 loopLibravatar Linus Torvalds1-12/+16
This helps a teeny bit. But what I -really- want to do is to avoid the whole 80-array loop, and do the xor updates as I go along.. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: minor fixupsLibravatar Junio C Hamano1-2/+2
Bert Wesarg noticed non-x86 version of SHA_ROT() had a typo. Also spell in-line assembly as __asm__(), otherwise I seem to get error: implicit declaration of function 'asm' from my compiler. Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: try to use rol/ror appropriatelyLibravatar Linus Torvalds1-10/+22
Use the one with the smaller constant. It _can_ generate slightly smaller code (a constant of 1 is special), but perhaps more importantly it's possibly faster on any uarch that does a rotate with a loop. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-06block-sha1: undo ctx->size changeLibravatar Junio C Hamano1-3/+3
Undo the change I picked up from the mailing list discussion suggested by Nico, not because it is wrong, but it will be done at the end of the follow-up series. Signed-off-by: Junio C Hamano <gitster@pobox.com>
2009-08-05Add new optimized C 'block-sha1' routinesLibravatar Linus Torvalds2-0/+166
Based on the mozilla SHA1 routine, but doing the input data accesses a word at a time and with 'htonl()' instead of loading bytes and shifting. It requires an architecture that is ok with unaligned 32-bit loads and a fast htonl(). Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Junio C Hamano <gitster@pobox.com>