public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
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


  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).