> From: Kat <Kaz****@t-tec*****> > Date: Mon, 17 Dec 2018 16:48:21 +1030 > > I support legacy software that uses MinGW. I have been experiencing an > interesting issue on a specific computer. Only happens on that computer. > It runs Windows 7 64 bit. First, are you running mingw.org's MinGW, or the MinGW64 fork (which can produce 64-bit binaries)? If the latter, this is the wrong list to ask about this problem. > I send a command via WriteFile to a device. This works just fine. The > device then starts returning data in multiple packets which I read > asynchronously using ReadFile(). I can use big hefty sleeps to prevent > access to the buffer during input. The corruption still happens. > > On all computers running Windows 10 this works fine. > > On this specific computer, some of the first bytes of a packet (I've > never seen it happen to the first packet) becomes corrupted. > > **** > > The only thing I can think of is that the ReadFile call itself is > somehow corrupting the data as it comes in (sometimes). Is this possible? I think you didn't tell enough details to answer that question. The code that opens the file/device and reads from it, as well as a few examples of the corruption would be the minimum information to begin thinking about this. Perhaps also consider to describe your asynchronous reading arrangement in more detail. > Since I use MinGW, I thought I would ask here first. Actually, ReadFile and WriteFile are _not_ MinGW, they are Win32 API functions implemented by Windows itself.
Fork