Bug 2167288 (CVE-2023-0045)

Summary:	CVE-2023-0045 kernel: Bypassing Spectre-BTI User Space Mitigations
Product:	[Other] Security Response	Reporter:	Sandipan Roy <saroy>
Component:	vulnerability	Assignee:	Nobody <nobody>
Status:	NEW ---	QA Contact:
Severity:	low	Docs Contact:
Priority:	low
Version:	unspecified	CC:	acaringi, bhu, chwhite, crwood, ddepaula, debarbos, dfreiber, dvlasenk, ezulian, fhrbata, ggasparb, hartsjc, hkrzesin, jarod, jburrell, jfaracco, jferlan, jforbes, jlelli, joe.lawrence, jshortt, jstancek, jwyatt, kcarcia, kernel-mgr, kyoshida, lgoncalv, lleshchi, lzampier, nmurray, ptalbert, qzhao, rkeshri, rogbas, rvrbovsk, scweaver, tyberry, vkumar, vpolasek, walters, williams, wmealing
Target Milestone:	---	Keywords:	Security
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:	A flaw was found in the Linux kernel. This issue occurs due to a failure mitigating the Spectre-BTI attack (using the kernel API), as IBPB is not issued during the syscall until the next schedule, leaving the system vulnerable.	Story Points:	---
Clone Of:		Environment:
Last Closed:		Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:	2167418, 2167419, 2167420, 2167421
Bug Blocks:	2167265

Description Sandipan Roy 2023-02-06 06:25:57 UTC

While testing the success rate of Spectre-BTI attacks, It was observed that the kernel does not issue an IBPB immediately during the syscall. The ib_prctl_set function updates the Thread Information Flags (TIFs) for the task and updates the SPEC_CTRL MSR on the function __speculation_ctrl_update, but the IBPB is only issued on the next schedule, when the TIF bits are checked. 

This leaves the victim vulnerable to values already injected on the BTB, prior to the prctl syscall. The behavior is only corrected after a reschedule of the taks happens. Furthermore, the kernel entrance (due to the syscall itself), does not issue an IBPB in the default scenarios (i.e., when the kernel protects itself via retpoline or eIBRS).

Refer:
https://github.com/es0j/CVE-2023-0045/blob/main/README.md
https://github.com/google/security-research/security/advisories/GHSA-9x5g-vmxf-4qj8