On Tue, Apr 20, 2021 at 09:30:08PM +0000, Michael Kelley wrote:

From: Joseph Salisbury <redacted> Sent: Friday, April 16, 2021 5:43 PM

quoted

There is not a consistent pattern for checking Hyper-V hypercall status.
Existing code uses a number of variants.  The variants work, but a consistent
pattern would improve the readability of the code, and be more conformant
to what the Hyper-V TLFS says about hypercall status.

Implemented new helper functions hv_result(), hv_result_success(), and
hv_repcomp().  Changed the places where hv_do_hypercall() and related variants
are used to use the helper functions.

Signed-off-by: Joseph Salisbury <redacted>
---
 arch/x86/hyperv/hv_apic.c           | 16 +++++++++-------
 arch/x86/hyperv/hv_init.c           |  2 +-
 arch/x86/hyperv/hv_proc.c           | 25 ++++++++++---------------
 arch/x86/hyperv/irqdomain.c         |  6 +++---
 arch/x86/hyperv/mmu.c               |  8 ++++----
 arch/x86/hyperv/nested.c            |  8 ++++----
 arch/x86/include/asm/mshyperv.h     |  1 +
 drivers/hv/hv.c                     |  2 +-
 drivers/pci/controller/pci-hyperv.c |  2 +-
 include/asm-generic/mshyperv.h      | 25 ++++++++++++++++++++-----
 10 files changed, 54 insertions(+), 41 deletions(-)

Wei -- I have gone through these changes reasonably carefully, but
your review of the Linux-in-the-root-partition changes would be a good
double-check.  It's too easy to get the sense of the tests backwards. :-(

The code looks correct to me.

Reviewed-by: Michael Kelley <redacted>

Thanks. I will apply this series to hyperv-next shortly.

Wei.