Jump to content

IT razno.


Mikorist

Preporučeni Komentari

9 minutes ago, shonne said:

glory, glory AMD :)

Pošto sam predvideo meltdown i spectre još prošle godine sam se prebacio na AMD ;)

Zjbvm se, de sam predvideo :D Nakon mislim 15 godina arčenja raznih Intela prešao na AMD zbog brezobrazne cene - u ceni poštenog i5 kupio gotovo celu AMD konfiguraciju sa FX8. Pa ti vidi. Nisam primetio nekakve zastoje, Photoshop, koderski programi, Corel, Illustrator i ponekad video editing, peva sve.

Jedino kuler mora, ko za Hiragu ( :D ), al jbg, šta sad.

Link to comment
Podeli na ovim sajtovima

5 minutes ago, Alex said:

Pošto sam predvideo meltdown i spectre još prošle godine sam se prebacio na AMD ;)

Zjbvm se, de sam predvideo :D Nakon mislim 15 godina arčenja raznih Intela prešao na AMD zbog brezobrazne cene - u ceni poštenog i5 kupio gotovo celu AMD konfiguraciju sa FX8. Pa ti vidi. Nisam primetio nekakve zastoje, Photoshop, koderski programi, Corel, Illustrator i ponekad video editing, peva sve.

Jedino kuler mora, ko za Hiragu ( :D ), al jbg, šta sad.

Da, FX je pristojan za rad za skoro sve stvari.

https://noctua.at/en/products/discontinued-products/nh-c12p-se14

14, ne cujem da li radi...800 o/min...

 

Link to comment
Podeli na ovim sajtovima

MarkJurich. github
 
 
 
 
Quote

 

To compile/link/run under Win7x64 & Visual Studio 2010, create a C++ Win32 Console Project (uncheck pre-compiled header check box), save as a .c source file, stay under "Debug" (no code optimization) vs "Release", and be aware of the following source code changes, depending if you start with this source or the original spectre.pdf research paper source (Copy & Pasted):
 
Legend: This Source Line Number(Original PDF Paper Source Line Number): OriginalStringSegment --> NewStringSegment // Comment
 
 
 

--( 13): */ --> */ // Copy & Paste from the Original PDF Source might need fixup at the end of comment
50( --): #define CACHE_HIT_THRESHOLD(80) --> #define CACHE_HIT_THRESHOLD (80) // There must be some whitespace between THRESHOLD and (80)
55( 39): int tries, i, j, k, mix_i, junk = 0; --> int tries, i, j, k, mix_i = 0; unsigned int junk = 0; // integer junk must be defined as unsigned integer junk
58( 42): volatile uint8_t * addr; --> volatile uint8_t * addr; volatile int z; // z must be defined at the beginning of the coding block
72( 56): volatile int z = 0; --> z = 0; // z declaration has been moved to the beginning of the coding block (above)
125(107): sscanf --> sscanf_s // to alleviate a compiler warning, sscanf must be sscanf_s
127(109): sscanf --> sscanf_s // to alleviate a compiler warning, sscanf must be sscanf_s
135(117): ’%c --> \'%c\' // fancy single quotes should be changed normal single quotes for command window
// displaying, but need to be escaped with backslashes in the double quoted string
136(118): "?" --> '?' ( ’?’ --> '?' ) // double or fancy quotes need to be normal single quotes
 
The created .exe (I called it Spectre.exe) should run in a Command Window. I tested two up-to-date Win7x64 Systems, with the following CPUs, and Positive Results:
 
Intel(R) Core(TM) i5-3320M CPU @ 2.60GHz (Ivy Bridge) lenovo X230
Intel(R) Core(TM) i7-3720QM CPU @ 2.60GHz (Ivy Bridge) lenovo W530
 
Quote
 
The i5 had the (KAISER) Windows Update for MeltDown/Spectre, which doesn't do anything to stop this variant of Spectre. I'm still waiting for any possible firmware updates to apply to either system. On the i5 (yesterday), I played with the Cache Hit Threshold from about 25 up to 200 with Positive Results, if I recall correctly.
 
Spectre Attacks (in research paper) can be Inter-Application (not necessarily targeting the kernel directly). Windows Updates are not restricted to OS Fixes. Firmware Updates can address some forms of attack. Software Patching (especially Signature Identification) can address immediate concerns. Is this going to stop Spectre Attacks, completely? Absolutely not (as the research paper alludes to). New Processor Designs eventually should...
 
... I would never say this code is reading kernel data. I mentioned it attacks itself as a demonstration. What we do know at this time, is that to start guarding against these forms of attacks (short of a new processor design), there will be much CPU overhead, and that overhead will depend on your current processor. I've read reports that some machine slowdowns as much as 30% may occur, with such solutions as speculation fencing, and that is what we have to look forward to, in the near future...
 
Intra-Application Spectre Attacks address the core flaw in Processor Design. Let's hope we can work together to protect against Spectre until machines/boxes get replaced due to attrition, etc.
 

 

 

 

 

Link to comment
Podeli na ovim sajtovima

Kreiraj nalog ili se prijavi da daš komentar

Potrebno je da budeš član DiyAudio.rs-a da bi ostavio komentar

Kreiraj nalog

Prijavite se za novi nalog na DiyAudio.rs zajednici. Jednostavno je!

Registruj novi nalog

Prijavi se

Već imaš nalog? Prijavi se ovde

Prijavi se odmah
  • Članovi koji sada čitaju   0 članova

    • Nema registrovanih članova koji gledaju ovu stranicu
×
×
  • Kreiraj novo...