Revert "usb: xhci: tegra: Fix error check"
authorDan Carpenter <dan.carpenter@linaro.org>
Tue, 4 Jul 2023 14:08:27 +0000 (17:08 +0300)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 3 Aug 2023 08:24:11 +0000 (10:24 +0200)
commit9ae3d7941f8d0e21112d9d7772eae61deba28fb3
tree34ddd447c0992a37bfff34e2b8ce39e74a73d7b0
parentcf8203ea19ef3917fdf97a72684061c84930e496
Revert "usb: xhci: tegra: Fix error check"

commit 288b4fa1798e3637a9304c6e90a93d900e02369c upstream.

This reverts commit 18fc7c435be3f17ea26a21b2e2312fcb9088e01f.

The reverted commit was based on static analysis and a misunderstanding
of how PTR_ERR() and NULLs are supposed to work.  When a function
returns both pointer errors and NULL then normally the NULL means
"continue operating without a feature because it was deliberately
turned off".  The NULL should not be treated as a failure.  If a driver
cannot work when that feature is disabled then the KConfig should
enforce that the function cannot return NULL.  We should not need to
test for it.

In this code, the patch means that certain tegra_xusb_probe() will
fail if the firmware supports power-domains but CONFIG_PM is disabled.

Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Fixes: 18fc7c435be3 ("usb: xhci: tegra: Fix error check")
Cc: stable <stable@kernel.org>
Link: https://lore.kernel.org/r/8baace8d-fb4b-41a4-ad5f-848ae643a23b@moroto.mountain
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/usb/host/xhci-tegra.c