Re: [PATCH] Bad sed expression in bootstrap.bat
On Mon, 2022-12-26 at 09:34 -0500, Paul Smith wrote: > > If you do this with "echo", you don't need to worry about backslash > > munging, I think, since cmd.exe's "echo" is very simple-minded, and > > just blindly copies every character, including the quotes. > > Oh... OK I will give this a try :). I realized I must have already known this at some point and forgotten it, since I used this method for other sed scripts in bootstrap.bat. Anyway I pushed a fix to bootstrap.bat that hopefully avoids the issue. I was still not ever able to reproduce the original failure on the Windows systems I have access to so I can't test it directly.
Re: Crash in hash_free_items()
On Fri, 2022-12-30 at 11:34 +0100, Gisle Vanem wrote: > I'm using the very latest GNU-make built on Win-10 by myself. Thanks for this report. Can you please be more clear about what you mean by "the very latest"? Do you mean the latest release 4.4? Or do you mean the latest code of the head of the Git branch? If the latter can you specify which SHA? I don't think checking this count is the correct solution; if there's a crash here it probably means we have either an uninitialized value, or freeing memory twice. Any information you can provide about the crash would be helpful. Can you provide information on which line in hash_free_items() actually failed, and what kind of crash you got? A repro case would be ideal but if that's too difficult we can try to investigate remotely. Thanks!
