Yes, your expression fatally confuses the JIT optimizer. The generated machine code looks like this: if ((value == null || value == new string[0]) == false) 00000027 test esi,esi ; value == null? 00000029 je 00000075 0000002b xor edx,edx ; new string[0] 0000002d mov ecx,6D913BD2h 00000032 call FFD20BC8 00000037 cmp eax,esi ; (value == new … Read more
A spot check on the generated machine code in the Release build shows that the x86 jitter doesn’t optimize the cast away. You have to look at the big picture here though. You are assigning properties of a control. They have a ton of side-effects. In the case of DateTimePicker, the assignment results in a … Read more
These are buffer overrun protection methods, and have nothing to do with compiler optimisation. MSVC will (if you specify the /GS switch) push a security cookie onto the stack near the return address so that it can detect a common case of stack corruption. Stack corruption can either be caused by bad code along the … Read more
Although it’s not a report in the sense of aggregated information, you might try the -fdump-ipa-all option which makes gcc produce dump files which at least keep you from having to analyse assembler code on what happened. Regarding loop optimization the -fdump-rtl-loop2 option might be of interest. For details on all this please see the … Read more
The portion of the C++ standard that applies is ยง1.9 “Program execution”. It reads, in part: conforming implementations are required to emulate (only) the observable behavior of the abstract machine as explained below. … A conforming implementation executing a well-formed program shall produce the same observable behavior as one of the possible execution sequences of … Read more
It depends. With the current version of Visual Studio, the methods that implement lambdas are never public, but they’re not always private. A simple program to test some versions of lambdas: public class Program { public static void Main() { var program = new Program(); Try(“A”, program.A); Try(“B”, program.B); Try(“C”, program.C); Console.ReadKey(); } private static … Read more
For example, 2 * f, might simply increment the exponent of f by 1, saving some cycles. This simply isn’t true. First you have too many corner cases such as zero, infinity, Nan, and denormals. Then you have the performance issue. The misunderstanding is that incrementing the exponent is not faster than doing a multiplication. … Read more
The behaviour is undefined in all C++ versions before C++17. The simple reason is that the two sides of the assignment operator can be evaluated in any order: Assuming A[1] is evaluated first, you get an int& referring to the second element of A at that point. Then, the func() is evaluated, which can reallocate … Read more
gcc -O3 uses a cmov for the conditional, so it lengthens the loop-carried dependency chain to include a cmov (which is 2 uops and 2 cycles of latency on your Intel Sandybridge CPU, according to Agner Fog’s instruction tables. See also the x86 tag wiki). This is one of the cases where cmov sucks. If … Read more
This is a code optimizer bug. It inlines both getData() and SetBit(). The combination appears to be fatal, it loses track of the value of 1 << ((pos) & 7) and always produces zero. This bug does not occur on VS2012. A workaround is to force one of the functions to not get inlined. Given … Read more