| Message ID | 1447795019-30176-19-git-send-email-ynorov@caviumnetworks.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi, Yury On 2015/11/18 5:16, Yury Norov wrote: > From: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> > > To make life for tools (such as gdb) easier when dealing with ILP32 processes, > we report a proper subarchitecture for ILP32 in the ELF auxiliary vectors. I saw some ilp32 relative patches in binutils mailing list. Does gdb fully support ilp32? Regards Bamvor
> On Nov 18, 2015, at 12:11 AM, Zhangjian (Bamvor) <bamvor.zhangjian@huawei.com> wrote: > > Hi, Yury > >> On 2015/11/18 5:16, Yury Norov wrote: >> From: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> >> >> To make life for tools (such as gdb) easier when dealing with ILP32 processes, >> we report a proper subarchitecture for ILP32 in the ELF auxiliary vectors. > I saw some ilp32 relative patches in binutils mailing list. Does gdb > fully support ilp32? I have a patch set but I have not tested them with the latest kernel patch set yet. The branch is located in the binutils-gdb git is https://sourceware.org/git/?p=binutils-gdb.git;a=shortlog;h=refs/heads/users/pinskia/gdb-aarch64-ilp32 . I think it will mostly work except for core support might need to be changed slightly. Thanks, Andrew > > Regards > > Bamvor >
On Wed, Nov 18, 2015 at 03:21:05AM -0800, pinskia@gmail.com wrote: > > > > On Nov 18, 2015, at 12:11 AM, Zhangjian (Bamvor) <bamvor.zhangjian@huawei.com> wrote: > > > > Hi, Yury > > > >> On 2015/11/18 5:16, Yury Norov wrote: > >> From: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> > >> > >> To make life for tools (such as gdb) easier when dealing with ILP32 processes, > >> we report a proper subarchitecture for ILP32 in the ELF auxiliary vectors. > > I saw some ilp32 relative patches in binutils mailing list. Does gdb > > fully support ilp32? > > I have a patch set but I have not tested them with the latest kernel patch set yet. The branch is located in the binutils-gdb git is https://sourceware.org/git/?p=binutils-gdb.git;a=shortlog;h=refs/heads/users/pinskia/gdb-aarch64-ilp32 . > I think it will mostly work except for core support might need to be changed slightly. > > Thanks, > Andrew > > Yes, it works mostly. I can read and modify text/data, set breakpoints etc. > > Regards > > > > Bamvor > >
Yury, We had implemented gdb-support against the the earlier ILP32 incarnation. I’ll take a look with the latest patch-set and see whether these still apply. Best, Philipp. > On 18 Nov 2015, at 21:25, Yury Norov <ynorov@caviumnetworks.com> wrote: > > On Wed, Nov 18, 2015 at 03:21:05AM -0800, pinskia@gmail.com wrote: >> >> >>> On Nov 18, 2015, at 12:11 AM, Zhangjian (Bamvor) <bamvor.zhangjian@huawei.com> wrote: >>> >>> Hi, Yury >>> >>>> On 2015/11/18 5:16, Yury Norov wrote: >>>> From: Philipp Tomsich <philipp.tomsich@theobroma-systems.com> >>>> >>>> To make life for tools (such as gdb) easier when dealing with ILP32 processes, >>>> we report a proper subarchitecture for ILP32 in the ELF auxiliary vectors. >>> I saw some ilp32 relative patches in binutils mailing list. Does gdb >>> fully support ilp32? >> >> I have a patch set but I have not tested them with the latest kernel patch set yet. The branch is located in the binutils-gdb git is https://sourceware.org/git/?p=binutils-gdb.git;a=shortlog;h=refs/heads/users/pinskia/gdb-aarch64-ilp32 . >> I think it will mostly work except for core support might need to be changed slightly. >> >> Thanks, >> Andrew >>> > > Yes, it works mostly. I can read and modify text/data, set breakpoints > etc. > >>> Regards >>> >>> Bamvor >>>
diff --git a/arch/arm64/include/asm/elf.h b/arch/arm64/include/asm/elf.h index 8f13dac..3e79569 100644 --- a/arch/arm64/include/asm/elf.h +++ b/arch/arm64/include/asm/elf.h @@ -163,9 +163,9 @@ extern int arch_setup_additional_pages(struct linux_binprm *bprm, #ifdef CONFIG_COMPAT #ifdef __AARCH64EB__ -#define COMPAT_ELF_PLATFORM ("v8b") +#define COMPAT_ELF_PLATFORM (is_ilp32_compat_task() ? "aarch64_be:ilp32" : "v8b") #else -#define COMPAT_ELF_PLATFORM ("v8l") +#define COMPAT_ELF_PLATFORM (is_ilp32_compat_task() ? "aarch64:ilp32" : "v8l") #endif #define COMPAT_ELF_ET_DYN_BASE (2 * TASK_SIZE_32 / 3)