MHI v5.13 commits for stable backporting

MHI v5.13 commits for stable backporting

[Manivannan Sadhasivam]

Hi Greg,

As suggested by you in MHI v5.13 PR [1], please find the below commits that
got merged for v5.13 and should be backported to stable kernels:

683e77cadc83 ("bus: mhi: core: Fix shadow declarations")
5630c1009bd9 ("bus: mhi: pci_generic: Constify mhi_controller_config struct definitions")
ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use")
47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset")
70f7025c854c ("bus: mhi: core: remove redundant initialization of variables state and ee")
68731852f6e5 ("bus: mhi: core: Return EAGAIN if MHI ring is full")
4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior")
6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up")
8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries")
0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue")
0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue")

I'll make sure to tag stable list for future potential patches.

Thanks,
Mani

[1] https://lore.kernel.org/patchwork/patch/1411161/#1609631

Re: MHI v5.13 commits for stable backporting

[Greg KH]

On Fri, May 07, 2021 at 01:37:36PM +0530, Manivannan Sadhasivam wrote:
> Hi Greg,
> 
> As suggested by you in MHI v5.13 PR [1], please find the below commits that
> got merged for v5.13 and should be backported to stable kernels:
> 
> 683e77cadc83 ("bus: mhi: core: Fix shadow declarations")
> 5630c1009bd9 ("bus: mhi: pci_generic: Constify mhi_controller_config struct definitions")
> ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use")
> 47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset")
> 70f7025c854c ("bus: mhi: core: remove redundant initialization of variables state and ee")
> 68731852f6e5 ("bus: mhi: core: Return EAGAIN if MHI ring is full")
> 4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior")
> 6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up")
> 8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries")
> 0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue")
> 0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue")
> 
> I'll make sure to tag stable list for future potential patches.

That's a lot, are you sure they are all needed?  Constify?

What order should these be applied in and how far back should the
patches be backported?

thanks,

greg k-h

Re: MHI v5.13 commits for stable backporting

[Manivannan Sadhasivam]

On Fri, May 07, 2021 at 10:17:48AM +0200, Greg KH wrote:
> On Fri, May 07, 2021 at 01:37:36PM +0530, Manivannan Sadhasivam wrote:
> > Hi Greg,
> > 
> > As suggested by you in MHI v5.13 PR [1], please find the below commits that
> > got merged for v5.13 and should be backported to stable kernels:
> > 
> > 683e77cadc83 ("bus: mhi: core: Fix shadow declarations")
> > 5630c1009bd9 ("bus: mhi: pci_generic: Constify mhi_controller_config struct definitions")
> > ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use")
> > 47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset")
> > 70f7025c854c ("bus: mhi: core: remove redundant initialization of variables state and ee")
> > 68731852f6e5 ("bus: mhi: core: Return EAGAIN if MHI ring is full")
> > 4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior")
> > 6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up")
> > 8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries")
> > 0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue")
> > 0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue")
> > 
> > I'll make sure to tag stable list for future potential patches.
> 
> That's a lot, are you sure they are all needed?  Constify?
>

Filtered the patches now...

> What order should these be applied in and how far back should the
> patches be backported?
> 

Below is the order and stable kernel revisions:

6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up") #5.10
47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset") #5.10
ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use") #5.10
8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries") #5.11
0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue") #5.11
4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior") #5.12
0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue") #5.12

TBH I misunderstood that Sasha's bot will pick the patches based on Fixes tag
and the commit message, etc... and will apply them to respective stable kernels
based on trial and error. Probably I expected too much or I'm lazy :)

Thanks,
Mani

> thanks,
> 
> greg k-h

Re: MHI v5.13 commits for stable backporting

[Greg KH]

On Fri, May 07, 2021 at 03:45:08PM +0530, Manivannan Sadhasivam wrote:
> On Fri, May 07, 2021 at 10:17:48AM +0200, Greg KH wrote:
> > On Fri, May 07, 2021 at 01:37:36PM +0530, Manivannan Sadhasivam wrote:
> > > Hi Greg,
> > > 
> > > As suggested by you in MHI v5.13 PR [1], please find the below commits that
> > > got merged for v5.13 and should be backported to stable kernels:
> > > 
> > > 683e77cadc83 ("bus: mhi: core: Fix shadow declarations")
> > > 5630c1009bd9 ("bus: mhi: pci_generic: Constify mhi_controller_config struct definitions")
> > > ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use")
> > > 47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset")
> > > 70f7025c854c ("bus: mhi: core: remove redundant initialization of variables state and ee")
> > > 68731852f6e5 ("bus: mhi: core: Return EAGAIN if MHI ring is full")
> > > 4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior")
> > > 6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up")
> > > 8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries")
> > > 0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue")
> > > 0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue")
> > > 
> > > I'll make sure to tag stable list for future potential patches.
> > 
> > That's a lot, are you sure they are all needed?  Constify?
> >
> 
> Filtered the patches now...
> 
> > What order should these be applied in and how far back should the
> > patches be backported?
> > 
> 
> Below is the order and stable kernel revisions:
> 
> 6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up") #5.10
> 47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset") #5.10
> ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use") #5.10
> 8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries") #5.11
> 0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue") #5.11
> 4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior") #5.12
> 0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue") #5.12
> 
> TBH I misunderstood that Sasha's bot will pick the patches based on Fixes tag
> and the commit message, etc... and will apply them to respective stable kernels
> based on trial and error. Probably I expected too much or I'm lazy :)

Please read
    https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.

Yes, we have scripts for lazy maintainers/developers to pick up things
that they forget to tag properly, but please, if at all possible, do it
correctly to ensure that the commits you know you want backported, are
properly backported and if not, you will be notified.

thanks,

greg k-h

Re: MHI v5.13 commits for stable backporting

[Greg KH]

On Fri, May 07, 2021 at 03:45:08PM +0530, Manivannan Sadhasivam wrote:
> On Fri, May 07, 2021 at 10:17:48AM +0200, Greg KH wrote:
> > On Fri, May 07, 2021 at 01:37:36PM +0530, Manivannan Sadhasivam wrote:
> > > Hi Greg,
> > > 
> > > As suggested by you in MHI v5.13 PR [1], please find the below commits that
> > > got merged for v5.13 and should be backported to stable kernels:
> > > 
> > > 683e77cadc83 ("bus: mhi: core: Fix shadow declarations")
> > > 5630c1009bd9 ("bus: mhi: pci_generic: Constify mhi_controller_config struct definitions")
> > > ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use")
> > > 47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset")
> > > 70f7025c854c ("bus: mhi: core: remove redundant initialization of variables state and ee")
> > > 68731852f6e5 ("bus: mhi: core: Return EAGAIN if MHI ring is full")
> > > 4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior")
> > > 6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up")
> > > 8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries")
> > > 0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue")
> > > 0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue")
> > > 
> > > I'll make sure to tag stable list for future potential patches.
> > 
> > That's a lot, are you sure they are all needed?  Constify?
> >
> 
> Filtered the patches now...
> 
> > What order should these be applied in and how far back should the
> > patches be backported?
> > 
> 
> Below is the order and stable kernel revisions:
> 
> 6403298c58d4 ("bus: mhi: core: Fix check for syserr at power_up") #5.10
> 47705c084659 ("bus: mhi: core: Clear configuration from channel context during reset") #5.10
> ec32332df764 ("bus: mhi: core: Sanity check values from remote device before use") #5.10
> 8de5ad994143 ("bus: mhi: core: Add missing checks for MMIO register entries") #5.11
> 0fccbf0a3b69 ("bus: mhi: pci_generic: Remove WQ_MEM_RECLAIM flag from state workqueue") #5.11
> 4547a749be99 ("bus: mhi: core: Fix MHI runtime_pm behavior") #5.12
> 0ecc1c70dcd3 ("bus: mhi: core: Fix invalid error returning in mhi_queue") #5.12

All now queued up, thanks.

greg k-h