Hi,
On 2023/7/18 21:27, Dan Carpenter wrote:
Basically everything in this email was wrong to a kind of shocking
degree. For example, ignoring kmalloc() failure is a bug so the fixes
tag is definitely warranted. But then you called me "bare brained"
which seems like a personal attack
Sorry, that's a misunderstanding
Sorry for my broken English.
by "bare brain", I means that by using the brains(pure code review) only,
I conjure up this adjective from the word "bare metal".
When I reply you email, I lack a word to describe this.
I believe that many experts have this sort of ability,
they could create a patch by simply give a glimpse of the code.
because they know how does the code run at the very low level.
so I'm going to report this as a code
of conduct violation.
Sorry Dan, you are welcomed.
Please withdraw this report.
I don't know why are you angry.
Because our hardware is rare,
Originally, by using the words "bare brain", I means by "pure brain",
I means that you probably without a hardware platform to do verification.
I want to know that if you have tested your patch on the board.
Or, I want to probe that if you have our hardware.
I will consider to send you one if you are long time contributor and if
you are really interested.
I have a lot of boards, now I'm feel a little bit tired by developing
drivers for all of them.
Please withdraw that report, personal attack tend to continues(across)
to multiple thread.
Sorry for my broken English. I will improve my written skill in the long
term.
Thanks for you contribution.
regards,
dan carpenter
On Tue, Jul 18, 2023 at 08:32:02PM +0800, suijingfeng wrote:
Hi,
Thanks for the patch.
The commit title generally should be 'drm/looongson: Add a check for
lsdc_bo_create() errors'
not 'drm: loongson: xxxx'
On 2023/7/18 15:01, Dan Carpenter wrote:
The lsdc_bo_create() function can fail so add a check for that.
Fixes: f39db26c5428 ("drm: Add kms driver for loongson display controller")
Please drop this Fixes tag, because you patch just add a error handling.
Yes, the lsdc_bo_create() function can fail, but this is generally not
happen.
Even if the fail happened, your patch is not fixing the root problem.
I know that you create this patch with the bare brain,
I would like hear more practical usage or bugs of this driver.
And mention more in the commit message is preferred.
Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
---
drivers/gpu/drm/loongson/lsdc_ttm.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/gpu/drm/loongson/lsdc_ttm.c b/drivers/gpu/drm/loongson/lsdc_ttm.c
index bb0c8fd43a75..bf79dc55afa4 100644
--- a/drivers/gpu/drm/loongson/lsdc_ttm.c
+++ b/drivers/gpu/drm/loongson/lsdc_ttm.c
@@ -496,6 +496,8 @@ struct lsdc_bo *lsdc_bo_create_kernel_pinned(struct drm_device *ddev,
int ret;
lbo = lsdc_bo_create(ddev, domain, size, true, NULL, NULL);
+ if (IS_ERR(lbo))
+ return ERR_CAST(lbo);
ret = lsdc_bo_reserve(lbo);
if (unlikely(ret)) {