Re: [PATCH 5.16 00/37] 5.16.14-rc1 review

[Greg Kroah-Hartman]

On Wed, Mar 09, 2022 at 11:05:15AM -0800, Guenter Roeck wrote:
> On 3/9/22 08:00, Greg Kroah-Hartman wrote:
> > This is the start of the stable review cycle for the 5.16.14 release.
> > There are 37 patches in this series, all will be posted as a response
> > to this one.  If anyone has any issues with these being applied, please
> > let me know.
> > 
> > Responses should be made by Fri, 11 Mar 2022 15:58:48 +0000.
> > Anything received after that time might be too late.
> > 
> 
> Almost all arm builds, all branches from 4.9.y to 5.16.y:
> 
> Error log:
> arch/arm/common/secure_cntvoff.S: Assembler messages:
> arch/arm/common/secure_cntvoff.S:24: Error: co-processor register expected -- `mcr p15,0,r0,c7,r5,4'
> arch/arm/common/secure_cntvoff.S:27: Error: co-processor register expected -- `mcr p15,0,r0,c7,r5,4'
> arch/arm/common/secure_cntvoff.S:29: Error: co-processor register expected -- `mcr p15,0,r0,c7,r5,4'
> 
> bisect log:
> 
> # bad: [3416254dac79ea26e08dffde371ab1fd3130223c] Linux 5.16.14-rc1
> # good: [6273c309621c9dd61c9c3f6d63f5d56ee2d89c73] Linux 5.16.13
> git bisect start 'HEAD' 'v5.16.13'
> # bad: [4403d69931dbc17659845d2d710602bbe35b4398] KVM: arm64: Allow indirect vectors to be used without SPECTRE_V3A
> git bisect bad 4403d69931dbc17659845d2d710602bbe35b4398
> # good: [6f0cf3a1eb8b46a5d652a395ba25a59c32a86692] ARM: report Spectre v2 status through sysfs
> git bisect good 6f0cf3a1eb8b46a5d652a395ba25a59c32a86692
> # bad: [87e96a363eb4a62b65bb974a46d518a87153cd1c] arm64: add ID_AA64ISAR2_EL1 sys register
> git bisect bad 87e96a363eb4a62b65bb974a46d518a87153cd1c
> # good: [654f0a73f042662a36155a0cafa30db846ccb5a9] ARM: use LOADADDR() to get load address of sections
> git bisect good 654f0a73f042662a36155a0cafa30db846ccb5a9
> # bad: [91bdae56c40ee6de675fba6ac283311c92c437ce] ARM: include unprivileged BPF status in Spectre V2 reporting
> git bisect bad 91bdae56c40ee6de675fba6ac283311c92c437ce
> # bad: [95ff4cb3b696a581d6166f0d754771bf9af5e27b] ARM: Spectre-BHB workaround
> git bisect bad 95ff4cb3b696a581d6166f0d754771bf9af5e27b
> # first bad commit: [95ff4cb3b696a581d6166f0d754771bf9af5e27b] ARM: Spectre-BHB workaround
> 
> Guenter

Wow these are all broken, have I mentioned I hate working on patches in
embargos?

I'll look at these in the morning, it's odd that Linus's tree isn't
showing these same issue, but I might have messed up the arm backports
somehow.  I did test them on using the multi_v7_defconfig configuration
and they built for me locally with gcc.  I'll try to see what is
different here...

thanks,

greg k-h