[Crash-utility] Re: [Crash-utility][RESEND PATCH v2 01/10] Add LoongArch64 framework code support
On 2023/12/25 12:56, Ming Wang wrote:
Hi, Kazu
Thanks for your review. Sorry, I forgot to send to the mailing list. So send it again.
On 12/13/23 12:57, HAGIO KAZUHITO(萩尾 一仁) wrote:
> Hi Ming,
>
> sorry for the delay.
>
> On 2023/11/08 10:39, Ming Wang wrote:
>
>> @@ -512,6 +527,10 @@ get_current_configuration(struct supported_gdb_version
*sp)
>> (target_data.target != MIPS64))
>> arch_mismatch(sp);
>>
>> + if ((target_data.initial_gdb_target == LOONGARCH64) &&
>> + (target_data.target != LOONGARCH64))
>> + arch_mismatch(sp);
>> +
> Please add the "if (target_data.target == X86_64)" block same as
> RISCV64 and etc. otherwise implied (sticky) target does not work:
>
> $ make target=LOONGARCH64
> ...
> $ make # without target
>
> The initial build in this source tree was for the LOONGARCH64 architecture.
>
> In order to build a crash binary for the X86_64 architecture:
> 1. remove the gdb-10.2 subdirectory
> 2. perform a "make clean"
> 3. retry the build
>
> make: *** [Makefile:253: all] Error 1
> $
Ok, I will fix it in V3.
>> @@ -2227,6 +2233,8 @@ struct offset_table { /* stash of
commonly-used offsets */
>> long module_memory_size;
>> long irq_data_irq;
>> long zspage_huge;
>> + long task_struct_thread_reg01;
>> + long task_struct_thread_reg03;
> These are not used in this patch, could you move to 05/10?
Ok, I will fix it in V3.
>
>> diff --git a/gdb-10.2.patch b/gdb-10.2.patch
>> index d81030d..c28fc38 100644
> Could you add "gdb-10.2/bfd/elf-bfd.h" and
"gdb-10.2/gdb/objfiles.h"
> to the list of the restored files at the beginning of gdb-10.2.patch?
> Please see the comment there for details. Now they each have two fixes.
I'm very sorry that I don't understand how to modify it here. Does it mean to
merge the modifications of the same file together?
No, sorry for lack of explanation. More specifically, please add this
patch:
--- a/gdb-10.2.patch
+++ b/gdb-10.2.patch
@@ -14,7 +14,9 @@ tar xvzmf gdb-10.2.tar.gz \
gdb-10.2/gdb/symfile.c \
gdb-10.2/gdb/Makefile.in \
gdb-10.2/gdb/dwarf2/read.c \
- gdb-10.2/gdb/ada-lang.c
+ gdb-10.2/gdb/ada-lang.c \
+ gdb-10.2/gdb/objfiles.h \
+ gdb-10.2/bfd/elf-bfd.h
exit 0
Here is the description of the tar command:
$ head -n 20 gdb-10.2.patch
# When this file is updated in an existing source tree, it gets re-applied
# during the next build using "patch -N --fuzz=0", which ignores patches
# that have already been applied. However, if a gdb file has been modified
# multiple times, the subsequent patching may fail to recognize that a
# given patch has been previously applied, and will attempt to re-apply it.
# To prevent any unintended consequences, this file also acts as a
# shell script that can restore any gdb file to its original state prior
# to all subsequent patch applications.
tar xvzmf gdb-10.2.tar.gz \
gdb-10.2/gdb/symtab.c \
gdb-10.2/gdb/printcmd.c \
gdb-10.2/gdb/symfile.c \
gdb-10.2/gdb/Makefile.in \
gdb-10.2/gdb/dwarf2/read.c \
gdb-10.2/gdb/ada-lang.c
exit 0
These files are modified multiple times with your patch set, so please
add the two files above.
$ grep ^+++ gdb-10.2.patch | sort | uniq -c | awk '$1 > 1'
2 +++ gdb-10.2/bfd/elf-bfd.h
3 +++ gdb-10.2/gdb/Makefile.in
2 +++ gdb-10.2/gdb/ada-lang.c
2 +++ gdb-10.2/gdb/dwarf2/read.c
2 +++ gdb-10.2/gdb/objfiles.h
2 +++ gdb-10.2/gdb/printcmd.c
3 +++ gdb-10.2/gdb/symfile.c
5 +++ gdb-10.2/gdb/symtab.c
Thanks,
Kazu