From: Luis Machado <luis.machado@arm.com>
To: Tomas Vanek <vanekt@fbl.cz>, gdb-patches@sourceware.org
Subject: Re: [RFC PATCH 1/5] gdb/arm: Introduce control_s and control_ns registers
Date: Wed, 9 Nov 2022 15:27:30 +0000 [thread overview]
Message-ID: <548988bd-5ccb-2982-296f-7eaa50328541@arm.com> (raw)
In-Reply-To: <6bdb8d6d-c0f7-c154-c0a9-4897759fce91@fbl.cz>
Hi,
On 11/8/22 15:58, Tomas Vanek wrote:
> On 08/11/2022 16:52, Tomas Vanek wrote:
>> Hi Luis,
>>
>> On 08/11/2022 12:23, Luis Machado wrote:
>>> Hi Tomas,
>>>
>>> On 11/5/22 09:44, Tomas Vanek wrote:
>>>> The M-profile security extension registers will be used for stack
>>>> selection in FNC_RETURN and return from secure to non-secure state.
>>>>
>>>> The presence of the registers is optional.
>>>> If one or both are missing the security extension profile is accepted.
>>>> The code using the registers must check m_profile_control_[n]?s_regnum
>>>> for -1.
>>>>
>>>> Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
>>>> ---
>>>> gdb/arm-tdep.c | 20 ++++++++++++++++++++
>>>> gdb/arm-tdep.h | 2 ++
>>>> 2 files changed, 22 insertions(+)
>>>>
>>>> diff --git a/gdb/arm-tdep.c b/gdb/arm-tdep.c
>>>> index 3105543..564ee43 100644
>>>> --- a/gdb/arm-tdep.c
>>>> +++ b/gdb/arm-tdep.c
>>>> @@ -9996,6 +9996,8 @@ enum arm_vfp_cprc_base_type
>>>> int m_profile_psp_ns_regnum = -1;
>>>> int m_profile_msp_s_regnum = -1;
>>>> int m_profile_psp_s_regnum = -1;
>>>> + int m_profile_control_s_regnum = -1;
>>>> + int m_profile_control_ns_regnum = -1;
>>>> int tls_regnum = 0;
>>>> /* If we have an object to base this architecture on, try to determine
>>>> @@ -10473,6 +10475,22 @@ enum arm_vfp_cprc_base_type
>>>> }
>>>> m_profile_psp_s_regnum = register_count++;
>>>> + /* Regard the control_s and control_ns registers optional,
>>>> + * accept the security extension feature even without them */
>>>> + valid_p = tdesc_numbered_register (feature, tdesc_data.get (),
>>>> + register_count, "control_s");
>>>
>>> What tools are advertising control_s and control_ns? We need to document those properly if we're planning on using them
>>> as part of decision-making in GDB.
>>
>> Good question.
>>
>> OpenOCD exposes them, like other M-profile security extension registers they have the feature name "v8-m.sp".
>> There is a pending patch to change the feature name to "org.gnu.gdb.arm.secext"
>> 7265: target/armv7m: fix feature name of ARMv8M security extension regs | https://review.openocd.org/c/openocd/+/7265
>>
>> PyOCD exposes them, like other M-profile security extension registers they have the feature name "v8-m.sp".
>
> Oops, PyOCD exposes just control, no control_s and control_ns variants.
>
Should we coordinate this a bit better between PyOCD and ST-LINK to expose the same set of registers. That would make it easier for
gdb to deal with it.
>>
>> ST-LINK GDB server. Version 5.6.0 (I assume not the latest version) exposes them (and other M-profile security extension registers too)
>> in the feature "org.gnu.gdb.arm.vfp"?? Obviously a bug!
>>
>> I have no access to Keil tools.
>>
>> Please be aware that control_s and control_ns are optional, the most of security extension support is functional
>> even in the case they are not accessible in the connected gdbserver.
>>
>>>
>>>
>>>> + if (!valid_p)
>>>> + warning (_("M-profile secext feature is missing required register control_s."));
>>>> + else
>>>> + m_profile_control_s_regnum = register_count++;
>>>> +
>>>> + valid_p = tdesc_numbered_register (feature, tdesc_data.get (),
>>>> + register_count, "control_ns");
>>>> + if (!valid_p)
>>>> + warning (_("M-profile secext feature is missing required register control_ns."));
>>>> + else
>>>> + m_profile_control_ns_regnum = register_count++;
>>>> +
>>>> have_sec_ext = true;
>>>> }
>>>> @@ -10553,6 +10571,8 @@ enum arm_vfp_cprc_base_type
>>>> tdep->m_profile_psp_ns_regnum = m_profile_psp_ns_regnum;
>>>> tdep->m_profile_msp_s_regnum = m_profile_msp_s_regnum;
>>>> tdep->m_profile_psp_s_regnum = m_profile_psp_s_regnum;
>>>> + tdep->m_profile_control_s_regnum = m_profile_control_s_regnum;
>>>> + tdep->m_profile_control_ns_regnum = m_profile_control_ns_regnum;
>>>> }
>>>> arm_register_g_packet_guesses (gdbarch);
>>>> diff --git a/gdb/arm-tdep.h b/gdb/arm-tdep.h
>>>> index bcd7e08..11670e9 100644
>>>> --- a/gdb/arm-tdep.h
>>>> +++ b/gdb/arm-tdep.h
>>>> @@ -133,6 +133,8 @@ struct arm_gdbarch_tdep : gdbarch_tdep_base
>>>> int m_profile_psp_ns_regnum = ARM_SP_REGNUM; /* M-profile PSP_NS register number. */
>>>> int m_profile_msp_s_regnum = ARM_SP_REGNUM; /* M-profile MSP_S register number. */
>>>> int m_profile_psp_s_regnum = ARM_SP_REGNUM; /* M-profile PSP_S register number. */
>>>> + int m_profile_control_s_regnum = -1; /* M-profile CONTROL_S register number. */
>>>> + int m_profile_control_ns_regnum = -1; /* M-profile CONTROL_NS register number. */
>>>> int tls_regnum = 0; /* Number of the tpidruro register. */
>>>
>>
>
next prev parent reply other threads:[~2022-11-09 15:27 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-05 9:44 Tomas Vanek
2022-11-05 9:44 ` [RFC PATCH 2/5] gdb/arm: PR 29716 Fix FNC_RETURN stack selection in exception unwinder Tomas Vanek
2022-11-11 9:23 ` Luis Machado
2022-11-05 9:44 ` [RFC PATCH 3/5] gdb/dwarf2: Add dwarf2_frame_reg_rule for GDB register number Tomas Vanek
2022-11-05 9:44 ` [RFC PATCH 4/5] gdb/arm: Unwinding of secure procedure with cmse_nonsecure_entry attribute Tomas Vanek
2022-11-05 9:44 ` [RFC PATCH 5/5] HACK frame inner than comparison for Arm M-profile sec ext Tomas Vanek
2022-11-08 11:23 ` [RFC PATCH 1/5] gdb/arm: Introduce control_s and control_ns registers Luis Machado
2022-11-08 15:52 ` Tomas Vanek
2022-11-08 15:58 ` Tomas Vanek
2022-11-09 15:27 ` Luis Machado [this message]
2022-11-08 16:48 ` Torbjorn SVENSSON
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=548988bd-5ccb-2982-296f-7eaa50328541@arm.com \
--to=luis.machado@arm.com \
--cc=gdb-patches@sourceware.org \
--cc=vanekt@fbl.cz \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).