On 1/19/23 03:15, Jan Kratochvil wrote:
> On Wed, 18 Jan 2023 22:22:51 +0800, Luis Machado wrote:
>> I'm on Ubuntu 22.04/20.04, but running 32-bit on a 64-bit system via Docker.
> 
> Ubuntu is the host OS. But I haven't found recent armv7 (32-bit) Ubuntu.
> What is the Docker container you use?

It is a custom one built with Ubuntu 22.04/20.04 images/packages.

> 
> 
>> x/i $pc
>> => 0xf7eadb04 <__libc_do_syscall+4>:    svc     0
>>
>> The test seems to be expecting a pattern of
>>
>> x/i $pc
>> => 0xf7eadb04 <__libc_do_syscall+4>:    svc     0x00000000
>>
>> If I replaced the pattern to force a match, the test continues until the unwinder check.
> 
> I forgot to update this pattern but it would not help the testcase anyway.
> The purpose is to stop in the function pthread_cond_timedwait.
> Here it stops in function __libc_do_syscall.
> It would be helpful at least a backtrace but I would more like to reproduce
> your environment.

I've attached a gdb.log file. Hopefully that will bring some light. Please note I used
"nosharedlibrary" to remove the symbols for my test.

> 
> 
>> Then it runs into a
>> UNTESTED again because I see the dwarf2 unwinder being used.
>>
>> That may be because I have debug info installed I suppose. I wonder if there
>> is a way to force gdb not to use the dwarf2 unwinder for libc, even if it is
>> available on the system.
> 
> I have fixed that part:
> 	gdb_test_no_output "set debug-file-directory" ""

For my case, it looks like gdb will still find the symbols. The command "nosharedlibrary" will
drop the symbols completely though. But then I see the prologue unwinders being used instead of the
exception frame.

> 
> 
> Thanks,
> Jan