x86/entry: Restore traditional SYSENTER calling convention
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 8 lines · tip-bot for Andy Lutomirski
Commit-ID: 30bfa7b3488bfb1bb75c9f50a5fcac1832970c60 Gitweb: http://git.kernel.org/tip/30bfa7b3488bfb1bb75c9f50a5fcac1832970c60 Author: Andy Lutomirski <[email protected]> AuthorDate: Wed, 16 Dec 2015 23:18:48 -0800 Committer: Thomas Gleixner <[email protected]> CommitDate: Mon, 21 Dec 2015 16:05:01 +0100x86/entry: Restore traditional SYSENTER calling convention
This is still missing from Linus' tree AFAICT, and I'm a bit nervous about it landing in 4.4 without ever showing up in a -rc.
--Andy
Quoted 9 lines · Andy Lutomirski
Quoted 7 lines · tip-bot for Andy Lutomirski
Commit-ID: 30bfa7b3488bfb1bb75c9f50a5fcac1832970c60 Gitweb: http://git.kernel.org/tip/30bfa7b3488bfb1bb75c9f50a5fcac1832970c60 Author: Andy Lutomirski <[email protected]> AuthorDate: Wed, 16 Dec 2015 23:18:48 -0800 Committer: Thomas Gleixner <[email protected]> CommitDate: Mon, 21 Dec 2015 16:05:01 +0100
Quoted 5 lines
Quoted 1 line
x86/entry: Restore traditional SYSENTER calling convention
This is still missing from Linus' tree AFAICT, and I'm a bit nervous about it landing in 4.4 without ever showing up in a -rc.
Linus has frequently stated that if it is something that is critical enough for stable, it is critical enough for final. Linus will decide if an additional -rc is needed for that reason.
-hpa
Quoted 12 lines · H. Peter Anvin
Quoted 11 lines · Andy Lutomirski
Quoted 8 lines · tip-bot for Andy Lutomirski
Commit-ID: 30bfa7b3488bfb1bb75c9f50a5fcac1832970c60 Gitweb: http://git.kernel.org/tip/30bfa7b3488bfb1bb75c9f50a5fcac1832970c60 Author: Andy Lutomirski <[email protected]> AuthorDate: Wed, 16 Dec 2015 23:18:48 -0800 Committer: Thomas Gleixner <[email protected]> CommitDate: Mon, 21 Dec 2015 16:05:01 +0100x86/entry: Restore traditional SYSENTER calling convention
Quoted 7 lines
Quoted 3 lines
This is still missing from Linus' tree AFAICT, and I'm a bit nervous about it landing in 4.4 without ever showing up in a -rc.
Linus has frequently stated that if it is something that is critical enough for stable, it is critical enough for final. Linus will decide if an additional -rc is needed for that reason.
True. It's also been in -next for a while, although the number of 32-bit -next testers over the holidays might not be that high.
--Andy
Quoted 4 lines · H. Peter Anvin
Linus has frequently stated that if it is something that is critical enough for stable, it is critical enough for final. Linus will decide if an additional -rc is needed for that reason.
So it would have been good to have it in an -rc, but at the same time I'm not particularly worried about this one.
It's not like it's complicated, and I'm assuming it got tested and passed all our current test-cases (which are much more complete than anything we've ever had historically).
Linus
Hi Linus and Andy,
We did test in our side with v4.4-rc8 + Andy's vDSO v2 patches + Android M (bionic libc using sysenter) ==> Device can boot up successfully
Other tests were:
- Android L (bionic libc using int80) + v4.4-rc8 ==> Device can boot up successfully
- Android L (bionic libc using int80) + v4.4-rc8 + Andy's v2 patches ==> Device can boot up successfully
- Android M (bionic libc using sysenter) + v4.4-rc8 ==> Device can NOT boot up successfully
- Android M (bionic libc using sysenter) + v4.4-rc8 + Andy's v2 patches ==> Device can boot up successfully
Thanks!
BR
qiuxu
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Linus Torvalds
Sent: Tuesday, January 5, 2016 3:28 AM
To: H. Peter Anvin
Cc: Andy Lutomirski; Shi, Mingwei; Fu, Borun; Gross, Mark; Andrew Lutomirski; Su, Tao; Borislav Petkov; Ingo Molnar; Brian Gerst; [email protected]; Zhuo, Qiuxu; Thomas Gleixner; Denys Vlasenko; Wang, Frank; [email protected]
Subject: Re: [tip:x86/urgent] x86/entry: Restore traditional SYSENTER calling convention
Quoted 4 lines · H. Peter Anvin
Linus has frequently stated that if it is something that is critical enough for stable, it is critical enough for final. Linus will decide if an additional -rc is needed for that reason.
So it would have been good to have it in an -rc, but at the same time I'm not particularly worried about this one.
It's not like it's complicated, and I'm assuming it got tested and passed all our current test-cases (which are much more complete than anything we've ever had historically).
Linus
���{.n�+�������+%�����ݶ��w��{.n�+����{��G�����{ay� ʇڙ�,j��f���h��������z_�(�階�ݢj"���m������G������?����&���~��iO��z��v�^�m���� ������?�I�