1
I Use This!
Very High Activity

Commits : Listings

Analyzed about 2 hours ago. based on code collected about 3 hours ago.
Aug 25, 2024 — Aug 25, 2025
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
core: new TTextWriter.AddHtmlAsText method More... 7 days ago
core: new TTextWriter.AddHtmlUnescape method More... 7 days ago
crypt: some AesNiEncryptCtrNist32By4() i386 asm tuning More... 7 days ago
crypt: updated documentation More... 7 days ago
crypt: TAesCtrOsl is not faster on i386 any more More... 8 days ago
crypt: experimental AES-CTR alternative code on x86_64 - our naive/simple x4 approach - which did wonders on i386 - is a bit slower on 86_64 for no obvious reason - perhaps newer CPUs may benefit from it - we also tried an x8 unrolled version to leverage the additional xmm8-xmm15 registers available on the platform - but it was actually slower than our x4 code - "simple" x4 and x8 code has been less tested than original AesNiEncryptCtrNist32 - we continue to use our original AesNiEncryptCtrNist32, and keep our naive/simple x4 and x8 code as disabled reference, and may retry with future CPUs More... 8 days ago
crypt: changed AesNiEncryptXmm0 usage - rax now expect TAesContext + $70 address natively More... 8 days ago
crypt: use Core i5 13400 numbers for AES documentation - and including our latest i386 asm for CTR More... 8 days ago
Merge pull request #374 from danielkuettner/patch-3
ab
More... 9 days ago
crypt: include some AesNiEncryptCtrNist32() numbers on Win32 More... 9 days ago
Update mormot.crypt.core.pas
DK
More... 9 days ago
crypt: introducing experimental AesNiEncryptCtrNist32By4() - only tested in POSIX, not on Win64 ABI - almost as fast as the previous x8 code - we will see if you could do better More... 11 days ago
tests: refined TTestCoreCrypto._AES internal steps More... 11 days ago
crypt: small gf_mul_pclmulqdq() optimization on FPC/i386 More... 11 days ago
crypt: i386 asm AES-CTR with 4x AES-NI interleave factor and SSE 4.1 - our code seems cleaner than OpenSSL overcomplicated asm, and also much faster from our tests More... 11 days ago
crypt: introducing USEAESNICTR conditional More... 11 days ago
crypt: modified TAesContext.buf definition More... 11 days ago
core: made FloatToJsonNan() inline More... 11 days ago
tests: enhanced JSON double NaN coverage More... 12 days ago
core: fixed FloatToJsonNan() and optimized More... 12 days ago
core: simplified Utf8ToFloatNan() More... 12 days ago
rest: fixed TRestServerAuthenticationSspi - avoid exception on missing 'data=...' parameter in URI - parse the "Authorization" protocol and catch only "Negotiate", allowing e.g. fallback to the "Basic" auth on the same URI More... 12 days ago
rest: some minor TRestServerUriContext refactoring More... 12 days ago
rest: rewritten TRestServerAuthenticationSspi process - to support only Negotiate - since NTLM three-way is not used any more, we can get rid of the overcomplicated server-side handshake state machine - is now in synch with our Web server authentication More... 12 days ago
rest: refactored TRestServer.Uri for cleaner and faster process More... 12 days ago
rest: TRestServerAuthenticationDefault/None cleanup More... 12 days ago
rest: refactored TRestClientAuthenticationSspi - to only support NTLM More... 12 days ago
net: properly remove NTLM from HTTP client authentiation documentation - will only support Negotiate from now on More... 12 days ago
lib: SSPI/GSSAPI "Negotiate" constants cleanup More... 12 days ago
lib: removed unsafe and deprecated NTLM from Windows SSPI authentication - Windows SSPI authentication is now only via Kerberso More... 12 days ago