From: John Baldwin <jhb@FreeBSD.org>
To: Simon Marchi <simark@simark.ca>, gdb-patches@sourceware.org
Subject: Re: [PATCH v4 04/13] x86 nat: Add helper functions to save the XSAVE layout for the host.
Date: Mon, 10 Apr 2023 14:00:50 -0700 [thread overview]
Message-ID: <0b60c557-ac1e-28ef-4bd6-06064232371e@FreeBSD.org> (raw)
In-Reply-To: <68edcaf2-367c-231c-d3b9-4a5ddbc9463f@simark.ca>
On 4/6/23 12:40 PM, Simon Marchi wrote:
> On 3/17/23 21:08, John Baldwin wrote:
>> x86_xsave_length returns the total length of the XSAVE state area
>> standard format as queried from CPUID.
>>
>> x86_fetch_xsave_layout uses CPUID to query the offsets of XSAVE
>> extended regions from the running host. The total length of the XSAVE
>> state area can either be supplied the caller if known (e.g. from
>> FreeBSD's PT_GETXSTATEINFO) or it can be queried from the running host
>> using x86_xsave_length.
>> ---
>> gdb/nat/x86-xstate.c | 65 ++++++++++++++++++++++++++++++++++++++++++++
>> gdb/nat/x86-xstate.h | 35 ++++++++++++++++++++++++
>> 2 files changed, 100 insertions(+)
>> create mode 100644 gdb/nat/x86-xstate.c
>> create mode 100644 gdb/nat/x86-xstate.h
>>
>> diff --git a/gdb/nat/x86-xstate.c b/gdb/nat/x86-xstate.c
>> new file mode 100644
>> index 00000000000..7b42efe1266
>> --- /dev/null
>> +++ b/gdb/nat/x86-xstate.c
>> @@ -0,0 +1,65 @@
>> +/* x86 XSAVE extended state functions.
>> +
>> + Copyright (C) 2022-2023 Free Software Foundation, Inc.
>> +
>> + This file is part of GDB.
>> +
>> + This program is free software; you can redistribute it and/or modify
>> + it under the terms of the GNU General Public License as published by
>> + the Free Software Foundation; either version 3 of the License, or
>> + (at your option) any later version.
>> +
>> + This program is distributed in the hope that it will be useful,
>> + but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + GNU General Public License for more details.
>> +
>> + You should have received a copy of the GNU General Public License
>> + along with this program. If not, see <http://www.gnu.org/licenses/>. */
>> +
>> +#include "gdbsupport/common-defs.h"
>> +#include "gdbsupport/x86-xstate.h"
>> +#include "nat/x86-cpuid.h"
>> +#include "nat/x86-xstate.h"
>> +
>> +/* Fetch the offset of a specific XSAVE extended region. */
>> +
>> +static int
>> +xsave_feature_offset (uint64_t xcr0, int feature)
>> +{
>> + uint32_t ebx;
>> +
>> + if ((xcr0 & (1ULL << feature)) == 0)
>> + return 0;
>> +
>> + if (!x86_cpuid_count (0xd, feature, nullptr, &ebx, nullptr, nullptr))
>
> Does x86_cpuid_count return a count, as the name suggests (and not a
> boolean)? If so, it should be == 0. Otherwise, ignore that.
No, it returns a bool like x86_cpuid. Some CPUID leaves on x86 take a
sub-leaf where the main leaf is specified in EAX and the sub-leaf in ECX
when CPUID is invoked. The first CPUID leaves that used this documented
the ECX input as a count (I think the first ones were used for enumerating
cache topology info where ECX was basically an index). As a result, the
APIs for CPUID builtins in GCC, etc. all use the suffix "_count" for the
variants that take the ECX sub-leaf as an input.
>> +/* See x86-xstate.h. */
>> +
>> +void
>> +x86_fetch_xsave_layout (uint64_t xcr0, int len, x86_xsave_layout &layout)
>> +{
>> + layout.sizeof_xsave = len;
>> + layout.avx_offset = xsave_feature_offset(xcr0, X86_XSTATE_AVX_ID);
>> + layout.bndregs_offset = xsave_feature_offset(xcr0, X86_XSTATE_BNDREGS_ID);
>> + layout.bndcfg_offset = xsave_feature_offset(xcr0, X86_XSTATE_BNDCFG_ID);
>> + layout.k_offset = xsave_feature_offset(xcr0, X86_XSTATE_K_ID);
>> + layout.zmm_h_offset = xsave_feature_offset(xcr0, X86_XSTATE_ZMM_H_ID);
>> + layout.zmm_offset = xsave_feature_offset(xcr0, X86_XSTATE_ZMM_ID);
>> + layout.pkru_offset = xsave_feature_offset(xcr0, X86_XSTATE_PKRU_ID);
>
> Missing spaces.
Oops, fixed.
>> diff --git a/gdb/nat/x86-xstate.h b/gdb/nat/x86-xstate.h
>> new file mode 100644
>> index 00000000000..e9f5fb57e3a
>> --- /dev/null
>> +++ b/gdb/nat/x86-xstate.h
>> @@ -0,0 +1,35 @@
>> +/* x86 XSAVE extended state functions.
>> +
>> + Copyright (C) 2022-2023 Free Software Foundation, Inc.
>> +
>> + This file is part of GDB.
>> +
>> + This program is free software; you can redistribute it and/or modify
>> + it under the terms of the GNU General Public License as published by
>> + the Free Software Foundation; either version 3 of the License, or
>> + (at your option) any later version.
>> +
>> + This program is distributed in the hope that it will be useful,
>> + but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + GNU General Public License for more details.
>> +
>> + You should have received a copy of the GNU General Public License
>> + along with this program. If not, see <http://www.gnu.org/licenses/>. */
>> +
>> +#ifndef NAT_X86_XSTATE_H
>> +#define NAT_X86_XSTATE_H
>> +
>> +#include "gdbsupport/x86-xstate.h"
>> +
>> +/* Return the size of the XSAVE extended state fetched via CPUID. */
>> +
>> +int x86_xsave_length ();
>> +
>> +/* Save the size and offsets of the XSAVE extended regions for the
>> + running host in LAYOUT. Offsets of each of the enabled regions in
>> + XCR0 are fetched via CPUID. */
>> +
>> +void x86_fetch_xsave_layout (uint64_t xcr0, int len, x86_xsave_layout &layout);
>
> The function could maybe return an x86_xsave_layout.
Ah, fair enough.
--
John Baldwin
next prev parent reply other threads:[~2023-04-10 21:00 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-18 1:08 [PATCH v4 00/13] Handle variable XSAVE layouts John Baldwin
2023-03-18 1:08 ` [PATCH v4 01/13] x86: Add an x86_xsave_layout structure to handle " John Baldwin
2023-03-28 11:35 ` George, Jini Susan
2023-04-10 19:28 ` John Baldwin
2023-04-06 19:09 ` Simon Marchi
2023-04-10 20:03 ` John Baldwin
2023-04-11 1:19 ` Simon Marchi
2023-04-11 14:23 ` Pedro Alves
2023-04-11 16:02 ` John Baldwin
2023-03-18 1:08 ` [PATCH v4 02/13] core: Support fetching TARGET_OBJECT_X86_XSAVE_LAYOUT from architectures John Baldwin
2023-04-06 19:28 ` Simon Marchi
2023-04-10 20:42 ` John Baldwin
2023-04-11 1:49 ` Simon Marchi
2023-04-11 16:06 ` John Baldwin
2023-04-11 16:21 ` Simon Marchi
2023-04-11 23:59 ` John Baldwin
2023-03-18 1:08 ` [PATCH v4 03/13] nat/x86-cpuid.h: Add x86_cpuid_count wrapper around __get_cpuid_count John Baldwin
2023-04-06 19:33 ` Simon Marchi
2023-04-10 20:49 ` John Baldwin
2023-04-11 1:49 ` Simon Marchi
2023-03-18 1:08 ` [PATCH v4 04/13] x86 nat: Add helper functions to save the XSAVE layout for the host John Baldwin
2023-04-06 19:40 ` Simon Marchi
2023-04-10 21:00 ` John Baldwin [this message]
2023-04-11 1:51 ` Simon Marchi
2023-03-18 1:08 ` [PATCH v4 05/13] gdb: Update x86 FreeBSD architectures to support XSAVE layouts John Baldwin
2023-04-06 19:54 ` Simon Marchi
2023-04-10 21:02 ` John Baldwin
2023-04-11 1:55 ` Simon Marchi
2023-03-18 1:08 ` [PATCH v4 06/13] gdb: Support XSAVE layouts for the current host in the FreeBSD x86 targets John Baldwin
2023-04-06 20:18 ` Simon Marchi
2023-04-10 21:27 ` John Baldwin
2023-04-11 2:23 ` Simon Marchi
2023-04-11 16:19 ` John Baldwin
2023-04-11 16:46 ` Simon Marchi
2023-04-11 21:37 ` John Baldwin
2023-04-11 22:35 ` John Baldwin
2023-04-12 14:35 ` Simon Marchi
2023-03-18 1:08 ` [PATCH v4 07/13] gdb: Update x86 Linux architectures to support XSAVE layouts John Baldwin
2023-04-07 1:43 ` Simon Marchi
2023-04-10 21:29 ` John Baldwin
2023-03-18 1:09 ` [PATCH v4 08/13] gdb: Support XSAVE layouts for the current host in the Linux x86 targets John Baldwin
2023-04-07 1:54 ` Simon Marchi
2023-03-18 1:09 ` [PATCH v4 09/13] gdb: Use x86_xstate_layout to parse the XSAVE extended state area John Baldwin
2023-04-07 2:13 ` Simon Marchi
2023-04-10 21:40 ` John Baldwin
2023-03-18 1:09 ` [PATCH v4 10/13] gdbserver: Add a function to set the XSAVE mask and size John Baldwin
2023-04-12 15:08 ` Simon Marchi
2023-04-27 17:24 ` John Baldwin
2023-03-18 1:09 ` [PATCH v4 11/13] gdbserver: Refactor the legacy region within the xsave struct John Baldwin
2023-04-12 18:34 ` Simon Marchi
2023-04-27 19:51 ` John Baldwin
2023-03-18 1:09 ` [PATCH v4 12/13] gdbserver: Read offsets of the XSAVE extended region via CPUID John Baldwin
2023-04-11 14:46 ` Pedro Alves
2023-04-11 16:25 ` John Baldwin
2023-04-12 19:11 ` Simon Marchi
2023-04-12 21:07 ` John Baldwin
2023-04-13 15:07 ` Simon Marchi
2023-03-18 1:09 ` [PATCH v4 13/13] x86: Remove X86_XSTATE_SIZE and related constants John Baldwin
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=0b60c557-ac1e-28ef-4bd6-06064232371e@FreeBSD.org \
--to=jhb@freebsd.org \
--cc=gdb-patches@sourceware.org \
--cc=simark@simark.ca \
/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).