Re: Build failures / crashes in stable queue branches

[Joel Fernandes <joel@joelfernandes.org>]

On 7/18/23 21:16, Yu Kuai wrote:
> Hi,
> 
> 在 2023/07/19 1:14, Joel Fernandes 写道:
>> On Tue, Jul 18, 2023 at 10:45 AM Greg KH <gregkh@linuxfoundation.org> 
>> wrote:
>>>
>>> On Tue, Jul 18, 2023 at 09:52:45AM -0400, Joel Fernandes wrote:
>>>> On 7/16/23 10:30, Greg KH wrote:
>>>>> On Sun, Jul 16, 2023 at 11:20:33AM +0800, Yu Kuai wrote:
>>>>>> Hi,
>>>>>>
>>>>>> 在 2023/07/15 23:49, Joel Fernandes 写道:
>>>>>>> Hi Yu,
>>>>>>>
>>>>>>> On Fri, Jul 14, 2023 at 03:21:46AM -0700, Guenter Roeck wrote:
>>>>>>> [..]
>>>>>>>> ---------
>>>>>>>> 6.1.y:
>>>>>>>>
>>>>>>>> Build reference: v6.1.38-393-gb6386e7314b4
>>>>>>>> Compiler version: alpha-linux-gcc (GCC) 11.4.0
>>>>>>>> Assembler version: GNU assembler (GNU Binutils) 2.40
>>>>>>>>
>>>>>>>> Building alpha:allmodconfig ... failed
>>>>>>>> Building m68k:allmodconfig ... failed
>>>>>>>> --------------
>>>>>>>> Error log:
>>>>>>>> <stdin>:1517:2: warning: #warning syscall clone3 not implemented 
>>>>>>>> [-Wcpp]
>>>>>>>> In file included from block/genhd.c:28:
>>>>>>>> block/genhd.c: In function 'disk_release':
>>>>>>>> include/linux/blktrace_api.h:88:57: error: statement with no 
>>>>>>>> effect [-Werror=unused-value]
>>>>>>>>       88 | # define 
>>>>>>>> blk_trace_remove(q)                            (-ENOTTY)
>>>>>>>>          
>>>>>>>> |                                                         ^
>>>>>>>> block/genhd.c:1185:9: note: in expansion of macro 
>>>>>>>> 'blk_trace_remove'
>>>>>>>>     1185 |         blk_trace_remove(disk->queue);
>>>>>>>
>>>>>>> 6.1 stable is broken and gives build warning without:
>>>>>>>
>>>>>>> cbe7cff4a76b ("blktrace: use inline function for 
>>>>>>> blk_trace_remove() while blktrace is disabled")
>>>>>>>
>>>>>>> Could you please submit it to stable for 6.1? (I could have done 
>>>>>>> that but it
>>>>>>> looks like you already backported related patches so its best for 
>>>>>>> you to do
>>>>>>> it, thanks for your help!).
>>>>>>
>>>>>> Thanks for the notice, however, I'll suggest to revert this patch for
>>>>>> now, because there are follow up fixes that is not applied yet.
>>>>>
>>>>> Which specific patch should be dropped?
>>>>>
>>>>
>>>> Yu: Ping? ;-). Are you suggesting the original set be reverted, or Greg
>>>> apply the above fix? Let us please keep 6.1 stable unbroken. ;-)
>>>>
>>>> Apologies for my noise if the issue has already been resolved.
>>>
>>> I think it has been resolved, but testing against the latest -rc release
>>> I sent out yesterday would be appreciated.
>>
>> Great.  Sure, I am going to run it today.
> 
> Sorry about the trouble, currently this patch will cause that scsi host
> module can't be unloaded, and there are follow up fixes for this:
> 
> https://lore.kernel.org/all/20230621160111.1433521-1-yukuai1@huaweicloud.com/
> https://lore.kernel.org/all/20230705024001.177585-1-yukuai1@huaweicloud.com/
> 
> The second patch is not applied yet, either revert the original patch or
> apply the above fix to stable as well.

The latest 6.1 stable now builds so it is all good.

thanks,

  - Joel