Can netimgui support X86?

[sammyfreg]

Certainly.

The NetImgui Client sources (code added to user application) should compile without any problems.

The NetImgui Server sources should still be able to compile in x86, but I haven't tested lately. This is not as critical as the client code, since the server application is meant to run on Windows (or MacOS/Linux with some additional changes).

On Thu, Jun 16, 2022 at 3:25 PM xiaoniu @.***> wrote:

— Reply to this email directly, view it on GitHub https://github.com/sammyfreg/netImgui/issues/34, or unsubscribe https://github.com/notifications/unsubscribe-auth/AESBPYZQ53NEWFDAAZZDV4LVPLCF7ANCNFSM5Y5VDZRA . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[IULOVE]

The client can compile X86 but there will be problems with running. I don’t know if you have tried it

[sammyfreg]

I just tried the client sample : SampleBasic in clang win32 and msvc win32 build (those are the 32bits x86 build). They both were able to run and connect to the NetImguiServer compiled in 32 and 64bits.

[IULOVE]

msvc 2019 Windows 10 19044.1706

I built the SampleBasic and Netimguiserver of the MSVC X86, and the following situation occurred under the operation of Debug

Failure in Assert (ispointer ());

Release:

0x004623b5 (located in SampleBasic_release.exe) abnormalities: 0xc0000005: Reading location 0x0e1b0e34 occurred.

[sammyfreg]

Thank you, this helps a lot.

I will try this repo with the exact same VisualStudio / Windows version. I also tried with VS2019. Do you know which update you are on?

Once I can reproduce it, should not be difficult to fix.

On Fri, Jun 17, 2022 at 12:05 PM xiaoniu @.***> wrote:

msvc 2019 Windows 10 19044.1706

I built the SampleBasic and Netimguiserver of the MSVC X86, and the following situation occurred under the operation of Debug [image: 1] https://user-images.githubusercontent.com/14193276/174216259-55163c1d-5c3f-49cb-8d93-757389d00d70.png [image: 2] https://user-images.githubusercontent.com/14193276/174216270-415495e2-60c1-46fa-92a3-9f5343f3a3e0.png

Failure in Assert (ispointer ());

Release: [image: 3] https://user-images.githubusercontent.com/14193276/174216419-888c879a-1c9e-42d3-b267-4c2992cac7a2.png [image: 4] https://user-images.githubusercontent.com/14193276/174216442-ae695463-fb18-462c-8d51-75b7d28c0ab2.png

0x004623b5 (located in SampleBasic_release.exe) abnormalities: 0xc0000005: Reading location 0x0e1b0e34 occurred.

— Reply to this email directly, view it on GitHub https://github.com/sammyfreg/netImgui/issues/34#issuecomment-1158440432, or unsubscribe https://github.com/notifications/unsubscribe-auth/AESBPY4YMSGOBOH44MUFHC3VPPTQHANCNFSM5Y5VDZRA . You are receiving this because you commented.Message ID: @.***>

[sammyfreg]

I think I see what your issue is. I fixed it on the 'Dev' branch, but it hasn't been pushed to main in a new release yet.

The simple fix is to add 1 line of code, so the method looks like this :

template <typename TType>
void OffsetPointer<TType>::SetPtr(TType* pPointer)
{
    mOffset     = 0; // Remove 'offset flag' that can be left active on non 64bits pointer **** THIS IS IMPORTANT FOR THE 32bits BUG***
    mPointer    = pPointer;
}

Or you can use the version from the 'Dev' branch.

[IULOVE]

dev no problem, thanks for your answer

[sammyfreg]

I will check this later, by compiling one of the samples in x86 in the coming weeks. Problem should not be too big, probably something to do with uint64/size_t identifiers.

On Thu., Jun. 16, 2022, 18:37 xiaoniu @.***> wrote:

The client can compile X86 but there will be problems with running. I don’t know if you have tried it

— Reply to this email directly, view it on GitHub https://github.com/sammyfreg/netImgui/issues/34#issuecomment-1157447477, or unsubscribe https://github.com/notifications/unsubscribe-auth/AESBPYYEJUDFSFZBPX55ALTVPLYVBANCNFSM5Y5VDZRA . You are receiving this because you commented.Message ID: @.***>