[Bug 51862] New: dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

List overview All Threads

newer

older

[Bug 48776] New: GoToMeeting...

[Bug 51403] New: MySQL workbench...

WineHQ Bugzilla

11 Oct 2021 11 Oct '21

12:18 a.m.

https://bugs.winehq.org/show_bug.cgi?id=51862

Bug ID: 51862 Summary: dlls/ntdll/unix/system.c: In function ‘init_cpu_info’ Product: Wine Version: 6.18 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: ntdll Assignee: wine-bugs@winehq.org Reporter: version2013@protonmail.com Distribution: ---

Created attachment 70774 --> https://bugs.winehq.org/attachment.cgi?id=70774 log.txt

in distro:

# uname --kernel-release 3.14.56

# gcc --version 4.8.4

# ldd --version 2.19

config line: configure --prefix=/usr --sysconfdir=/etc --localstatedir=/var --with-x --libdir=/usr/lib32 CFLAGS="-O2 -march=i686 -mtune=i686"

Compiling wine-6.18 and wine-6.19, fails with:

/dlls/ntdll/unix/system.c: In function ‘init_cpu_info’: /dlls/ntdll/unix/system.c:250:5: error: inconsistent operand constraints in an ‘asm’

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

Show replies by date

WineHQ Bugzilla

11 Oct 11 Oct

12:19 a.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Nathan Bennett version2013@protonmail.com changed:

What |Removed |Added ---------------------------------------------------------------------------- Distribution|--- |Other

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

6:31 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Fabian Maurer dark.shadow4@web.de changed:

What |Removed |Added ---------------------------------------------------------------------------- CC| |dark.shadow4@web.de

--- Comment #1 from Fabian Maurer dark.shadow4@web.de --- Not sure what the minimum supported GCC version is, but is there a reason you can't update your GCC?

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

12 Oct 12 Oct

2:34 a.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #2 from Nathan Bennett version2013@protonmail.com --- Even if I could update GCC in this specific distro version I am using, I was hoping to continue using this older distro with newer versions of wine.

It appears I may have to cease compiling wine in this specific distro version. Last wine that compiled successfully was 6.16.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

3:19 a.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #3 from Nathan Bennett version2013@protonmail.com --- Clarifying what I wrote; updating my GCC is irrelevant to my point.

I was hoping to continue using older distros, with older GCC, to compile newer versions of wine.

I suppose it is to be expected, as wine is updated, another older GCC version fails to compile the updated wine.

A minimum GCC version is absent official documentation, as far as I am aware.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

5:45 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Fabian Maurer dark.shadow4@web.de changed:

--- Comment #4 from Fabian Maurer dark.shadow4@web.de --- Confirming.

Well, not entirely sure why you'd use that old of a distro, but I guess you have your reasons. I personally don't see why we shouldn't support that, as long as the effort isn't too much.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

5:47 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #5 from Fabian Maurer dark.shadow4@web.de --- Created attachment 70787 --> https://bugs.winehq.org/attachment.cgi?id=70787 Patch to fix compilation with gcc 4.8

Attaching a patch that *should* fix the issue. Not 100% sure, since I use a newer distro and running code compiled with gcc 4.8 crashes with a segfault instantly... Please report back on whether that fixes the issue for you, and whether wine actually works with that.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

5:50 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Zebediah Figura z.figura12@gmail.com changed:

What |Removed |Added ---------------------------------------------------------------------------- CC| |z.figura12@gmail.com

--- Comment #6 from Zebediah Figura z.figura12@gmail.com --- (In reply to Fabian Maurer from comment #5)

...

Created attachment 70787 [details] Patch to fix compilation with gcc 4.8

Attaching a patch that *should* fix the issue. Not 100% sure, since I use a newer distro and running code compiled with gcc 4.8 crashes with a segfault instantly... Please report back on whether that fixes the issue for you, and whether wine actually works with that.

That can't work. There's no guarantee that ebx doesn't get scratched between two separate asm blocks, and you're missing an ebx clobber in the first block anyway.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

11:24 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #7 from Fabian Maurer dark.shadow4@web.de --- Created attachment 70789 --> https://bugs.winehq.org/attachment.cgi?id=70789 Patch v2

Thanks, I knew I must've missed something... I mean, it worked, but I admit I don't know what you all need for inline assembly to *reliably* work.

Anyways, the point is that we can't clobber ebx. What about the new patch I attached?

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

11:24 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Fabian Maurer dark.shadow4@web.de changed:

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

14 Oct 14 Oct

3:26 a.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #8 from Nathan Bennett version2013@protonmail.com --- (In reply to Fabian Maurer from comment #7) Used Patch v2 Compiling finished with 'Wine Build Complete'.

Thank you.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

17 Nov 17 Nov

4:26 a.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #9 from Nathan Bennett version2013@protonmail.com --- (In reply to Fabian Maurer from comment #7) wine-6.21 Used Patch v2 Compiling finished with 'Wine Build Complete'.

Thank you.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

28 Nov 28 Nov

12:13 a.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

winetest@luukku.com changed:

What |Removed |Added ---------------------------------------------------------------------------- CC| |winetest@luukku.com

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

14 Dec 14 Dec

8:27 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Jacek Caban jacek@codeweavers.com changed:

What |Removed |Added ---------------------------------------------------------------------------- CC| |jacek@codeweavers.com

--- Comment #10 from Jacek Caban jacek@codeweavers.com --- Created attachment 71290 --> https://bugs.winehq.org/attachment.cgi?id=71290 fix

We could use something like the attached patch to avoid the problem. In theory we have a similar problem in __cpuidex from intrin.h. That one is used only for PE parts, so an archaic host GCC does not use it as long as mingw is available.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

10:54 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #11 from Fabian Maurer dark.shadow4@web.de --- @Jacek Caban Yes, that looks similar to how it was before it got changed. However, isn't my shorter version preferable?

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

15 Dec 15 Dec

2:41 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #12 from Jacek Caban jacek@codeweavers.com --- When inline assembly is causing problems, I think it's safer to just avoid inline assembly instead of working around problems.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

4:39 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #13 from Fabian Maurer dark.shadow4@web.de --- It's a well known issue with older gcc (can't clobber ebx), I don't see how that's unsafe. I mean, the x64 version already works, just need a tweak for 32bit. I would prefer the simpler version, since it's a lot less invasive. Just backup and restore the register.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

17 Dec 17 Dec

9:42 p.m.

New subject: [Bug 51862] dlls/ntdll/unix/system.c: In function ‘init_cpu_info’

https://bugs.winehq.org/show_bug.cgi?id=51862

Jacek Caban jacek@codeweavers.com changed:

--- Comment #14 from Jacek Caban jacek@codeweavers.com --- Inline assembly works around the problem while full assembly implementation avoids the problem in the first place and it's less invasive comparing to pre-regression code.

It's fixed in git.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

20 Dec 20 Dec

12:35 p.m.

New subject: [Bug 51862] Error compiling function init_cpu_info() in dlls/ntdll/unix/system.c with GCC 4.8.4

https://bugs.winehq.org/show_bug.cgi?id=51862

Matteo Bruni matteo.mystral@gmail.com changed:

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

21 Dec 21 Dec

6:44 a.m.

New subject: [Bug 51862] Error compiling function init_cpu_info() in dlls/ntdll/unix/system.c with GCC 4.8.4

https://bugs.winehq.org/show_bug.cgi?id=51862

--- Comment #15 from Nathan Bennett version2013@protonmail.com --- Compiling wine-7.0-rc2 results in "Wine build complete" Thank you.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

WineHQ Bugzilla

26 Dec 26 Dec

4:44 p.m.

New subject: [Bug 51862] Error compiling function init_cpu_info() in dlls/ntdll/unix/system.c with GCC 4.8.4

https://bugs.winehq.org/show_bug.cgi?id=51862

Alexandre Julliard julliard@winehq.org changed:

What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED

--- Comment #16 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 7.0-rc3.

-- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1116

Age (days ago)

1192

Last active (days ago)

wine-bugs@winehq.org

20 comments

1 participants

tags (0)

participants (1)

WineHQ Bugzilla