RE: [EXTERNAL] Re: Windows Application Issues | PostgreSQL | REF # 48475607

Thanks for the reply, Andrew and Thomas. The issue is originally from user feedback. And our team verified and reproduced it too. And it is high priority from our end is because: 1. It is regression, it did not see it in the previous build. 2. After the crash happened, even the app could be installed, but cluster initialization failed. And due to which we cannot connect to the database and proceed further.

@Shawn @Rahul, do you have any comment or suggestions?

Regards!
Haifang

-----Original Message-----
From: Andrew Dunstan <andrew(at)dunslane(dot)net>
Sent: Thursday, August 29, 2024 8:17 AM
To: Thomas Munro <thomas(dot)munro(at)gmail(dot)com>; Haifang Wang (Centific Technologies Inc) <v-haiwang(at)microsoft(dot)com>
Cc: Sandeep Thakkar <sandeep(dot)thakkar(at)enterprisedb(dot)com>; Rahul Pandey <pandeyrah(at)microsoft(dot)com>; Vishwa Deepak <Vishwa(dot)Deepak(at)microsoft(dot)com>; Shawn Steele <Shawn(dot)Steele(at)microsoft(dot)com>; Amy Wishnousky <amyw(at)microsoft(dot)com>; Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>; pgsql-bugs(at)lists(dot)postgresql(dot)org; Shweta Gulati <gulatishweta(at)microsoft(dot)com>; Ashish Nawal <nawalashish(at)microsoft(dot)com>
Subject: Re: [EXTERNAL] Re: Windows Application Issues | PostgreSQL | REF # 48475607

[You don't often get email from andrew(at)dunslane(dot)net(dot) Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]

On 2024-08-28 We 6:16 AM, Thomas Munro wrote:
> On Tue, Aug 27, 2024 at 9:29 AM Haifang Wang (Centific Technologies
> Inc) <v-haiwang(at)microsoft(dot)com> wrote:
>> Just would like to follow up with you about this bug. How is it going with the investigation?
> Well, we figured out that Windows is aborting inside setlocale(). The
> C standard says "The pointer to string returned by the setlocale
> function is such that a subsequent call with that string value and its
> associated category will restore that part of the program’s locale."
> Window does not conform, and crashes itself.
>
> One practical change we could make would be to reject non-ASCII locale
> names in initdb, as in the attached, and back-patch to 16. Then the
> EDB installer (a separate project that is calling initdb and providing
> the non-ASCII string containing "ü") would at least get a graceful
> failure instead of a crash. It hardly makes a difference to the user,
> but I've finally understood that these crashes in the Windows runtime
> library might be generating a stream of crash reports to Microsoft,
> which might be bringing you here?
>
> For example:
>
> $ initdb -D asdf --locale="liberté"
> The files belonging to this database system will be owned by user "tmunro".
> This user must also own the server process.
>
> initdb: error: locale name "liberté" contains non-ASCII characters

+1 for doing this.

cheers

andrew

--
Andrew Dunstan
EDB: https://www.enterprisedb.com/