Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-25 Thread Kirill A. Shutemov
On Wed, Oct 24, 2018 at 11:49:17AM -0700, Andy Lutomirski wrote: > On Tue, Oct 23, 2018 at 9:32 AM Kirill A. Shutemov > wrote: > > > > modify_ldt(2) leaves old LDT mapped after we switch over to the new one. > > Memory for the old LDT gets freed and the pages can be re-used. > > > > Leaving the

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-25 Thread Kirill A. Shutemov
On Wed, Oct 24, 2018 at 11:49:17AM -0700, Andy Lutomirski wrote: > On Tue, Oct 23, 2018 at 9:32 AM Kirill A. Shutemov > wrote: > > > > modify_ldt(2) leaves old LDT mapped after we switch over to the new one. > > Memory for the old LDT gets freed and the pages can be re-used. > > > > Leaving the

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-24 Thread H. Peter Anvin
On 10/23/18 9:31 AM, Kirill A. Shutemov wrote: > > It shouldn't be a particularly hot path anyway. > That's putting it mildly. -hpa

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-24 Thread H. Peter Anvin
On 10/23/18 9:31 AM, Kirill A. Shutemov wrote: > > It shouldn't be a particularly hot path anyway. > That's putting it mildly. -hpa

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-24 Thread Andy Lutomirski
On Tue, Oct 23, 2018 at 9:32 AM Kirill A. Shutemov wrote: > > modify_ldt(2) leaves old LDT mapped after we switch over to the new one. > Memory for the old LDT gets freed and the pages can be re-used. > > Leaving the mapping in place can have security implications. The mapping > is present in

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-24 Thread Andy Lutomirski
On Tue, Oct 23, 2018 at 9:32 AM Kirill A. Shutemov wrote: > > modify_ldt(2) leaves old LDT mapped after we switch over to the new one. > Memory for the old LDT gets freed and the pages can be re-used. > > Leaving the mapping in place can have security implications. The mapping > is present in

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-24 Thread Christoph Hellwig
The subject line does not parse..

Re: [PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-24 Thread Christoph Hellwig
The subject line does not parse..

[PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-23 Thread Kirill A. Shutemov
modify_ldt(2) leaves old LDT mapped after we switch over to the new one. Memory for the old LDT gets freed and the pages can be re-used. Leaving the mapping in place can have security implications. The mapping is present in userspace copy of page tables and Meltdown-like attack can read these

[PATCH 2/2] x86/ldt: Unmap PTEs for the slow before freeing LDT

2018-10-23 Thread Kirill A. Shutemov
modify_ldt(2) leaves old LDT mapped after we switch over to the new one. Memory for the old LDT gets freed and the pages can be re-used. Leaving the mapping in place can have security implications. The mapping is present in userspace copy of page tables and Meltdown-like attack can read these