On Tue, 19 Nov 2019, Jani Nikula <jani.nikula@xxxxxxxxxxxxxxx> wrote: > On Mon, 18 Nov 2019, Thierry Reding <thierry.reding@xxxxxxxxx> wrote: >> On Mon, Nov 18, 2019 at 07:46:10PM +0800, zhengbin wrote: >>> zhengbin (2): >>> drm/tegra: Use PTR_ERR_OR_ZERO() to simplify code in >>> tegra_bo_dumb_create >>> drm/tegra: Use PTR_ERR_OR_ZERO() to simplify code in tegra_gem_create >>> >>> drivers/gpu/drm/tegra/drm.c | 5 +---- >>> drivers/gpu/drm/tegra/gem.c | 5 +---- >>> 2 files changed, 2 insertions(+), 8 deletions(-) >> >> As I explained in response to the same patches sent for other drivers >> already, I don't think this has any merit. > > I agree completely. > > Apparently there's a coccicheck flagging constructs like this; perhaps > that should be addressed. Julia? > > Things like: > > drivers/gpu/drm/tegra/gem.c:457:1-3: WARNING: PTR_ERR_OR_ZERO can be used > > leading to: > > - if (IS_ERR(bo)) > - return PTR_ERR(bo); > - > - return 0; > + return PTR_ERR_OR_ZERO(bo); > > I think we have consensus the error path and the happy day scenarios > should remain distinct. Moreover, I find PTR_ERR_OR_ZERO() causes me to > pause for a moment while the original is a crystal clear immediately. Re consensus, [1][2] and probably more. Per an email search, this pops up for the same stuff over and over again. Can we please just remove the cocci check? BR, Jani. [1] http://mid.mail-archive.com/daade52d-0dfc-9365-b17c-02e7e785afad@xxxxxx [2] http://mid.mail-archive.com/20191118130022.GM4345@xxxxxxxxxxx -- Jani Nikula, Intel Open Source Graphics Center
