Re: [PATCH v10 00/11] Add TDX Guest Support (Initial support)

[Kuppuswamy, Sathyanarayanan]

On 10/9/21 12:38 AM, Borislav Petkov wrote:
> On Fri, Oct 08, 2021 at 10:37:36PM -0700, Kuppuswamy Sathyanarayanan wrote:
> > Hi All,
>
> Now let's see: you sent this particular patchset on Monday, 4th. The
> usual process is that you wait at least a week for review comments,
> incorporate them into your next revision and then you send it. We were
> still reviewing v8...

Sorry for the quick re-submissions. But the main reason for sending v9
within a week is,

1. Following compilation error found in v8.

This fails to build:

arch/x86/kernel/tdx.c: In function ‘tdx_write_msr_safe’:
arch/x86/kernel/tdx.c:135:22: error: implicit declaration of function ‘tdx_is_context_switched_msr’ 
[-Werror=implicit-function-declaration]
135 | WARN_ON_ONCE(tdx_is_context_switched_msr(msr));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
./include/asm-generic/bug.h:104:32: note: in definition of macro ‘WARN_ON_ONCE’
104 | int __ret_warn_on = !!(condition);


2. I had to rebase my patches on your tip tree again to adapt to the latest
version of CC patches.
3. Also to address your comment about using is_tdx_guest() in
cc_platform_has()

I thought the above issues warranted a re-submission. I know that it is a mistake
from my end. But I did not want you to review code changes that might go away
due to rebase.


> But I see already a v9 in my mbox from yesterday and *also* a v10. v9
> you probably didn't build-test enough so you had to hastily do a v10. 4
> days later!

I have sent v10 within few hours of v9 submission to fix a static inline issue.

I did not catch it my compilation test because, it happens only with a
TDX disabled config.

Sorry for the trouble again. Please ignore the v9 version. I will try not to repeat
this  in future.

> And because that's not enough, there are a bunch of other TDX patchsets
> from you flying in constantly.
>
> Now, please explain to me how you imagine this whole review thing is
> supposed to work?
>
> You hammer people with patchsets until they go in? Forget proper review?
>
> Or people should drop the other things they have to do for their jobs
> and deal only with your patchsets?
>
> How about we trade places: you review and try to get sh*t to work and I
> hammer you with patchsets every 3-4 days?
>
> For chrissakes, please calm down with that constant hammering and try to
> put yourself in the maintainer's shoes for once. Also, try to realize
> that hammering people with patchsets will get you the *opposite* of what
> you're trying to achieve - you will get ignored.
>
> Geez.

--
Sathyanarayanan Kuppuswamy
Linux Kernel Developer