All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dan Williams <dan.j.williams@intel.com>
To: bp@alien8.de
Cc: Borislav Petkov <bp@suse.de>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>,
	x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
	Tony Luck <tony.luck@intel.com>,
	linux-kernel@vger.kernel.org, linux-nvdimm@lists.01.org
Subject: [PATCH] x86/mce: Gate copy_mc_fragile() export by CONFIG_COPY_MC_TEST=y
Date: Wed, 07 Oct 2020 11:24:32 -0700	[thread overview]
Message-ID: <160209507277.2768223.9933672492157583642.stgit@dwillia2-desk3.amr.corp.intel.com> (raw)
In-Reply-To: <20201007111447.GA23257@zn.tnic>

It appears that modpost is not happy about exporting assembly symbols
that are not consumed in the same build. As Boris reports:

    WARNING: modpost: EXPORT symbol "copy_mc_fragile" [vmlinux] version generation failed, symbol will not be versioned.

The export is only consumed in the CONFIG_COPY_MC_TEST=y case, and even
then not in a way that modpost could see. CONFIG_COPY_MC_TEST uses a
module built in tools/testing/nvdimm/ to exercise the copy_mc_fragile()
corner cases.  Given the test already requires manually editing the
config entry for CONFIG_COPY_MC_TEST to make it "def_bool y" the
additional dependency to require is CONFIG_MODVERSIONS=n is not too
onerous.

Alternatively, COPY_MC_TEST and its related infrastructure could just be
ripped out because it has served its purpose. For now, just stop
exporting the symbol by default, and add the MODVERSIONS dependency to
the test.

Fixes: ec6347bb4339 ("x86, powerpc: Rename memcpy_mcsafe() to copy_mc_to_{user, kernel}()")
Reported-by: Borislav Petkov <bp@suse.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: x86@kernel.org
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Tony Luck <tony.luck@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 arch/x86/Kconfig.debug    |    1 +
 arch/x86/lib/copy_mc_64.S |    2 ++
 2 files changed, 3 insertions(+)

diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index 27b5e2bc6a01..6f0f5d8ac62e 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -63,6 +63,7 @@ config EARLY_PRINTK_USB_XDBC
 	  crashes or need a very simple printk logging facility.
 
 config COPY_MC_TEST
+	depends on !MODVERSIONS
 	def_bool n
 
 config EFI_PGT_DUMP
diff --git a/arch/x86/lib/copy_mc_64.S b/arch/x86/lib/copy_mc_64.S
index 892d8915f609..50fb05256751 100644
--- a/arch/x86/lib/copy_mc_64.S
+++ b/arch/x86/lib/copy_mc_64.S
@@ -88,7 +88,9 @@ SYM_FUNC_START(copy_mc_fragile)
 .L_done:
 	ret
 SYM_FUNC_END(copy_mc_fragile)
+#ifdef CONFIG_COPY_MC_TEST
 EXPORT_SYMBOL_GPL(copy_mc_fragile)
+#endif
 
 	.section .fixup, "ax"
 	/*
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

WARNING: multiple messages have this Message-ID (diff)
From: Dan Williams <dan.j.williams@intel.com>
To: bp@alien8.de
Cc: Borislav Petkov <bp@suse.de>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>,
	x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
	Tony Luck <tony.luck@intel.com>,
	linux-kernel@vger.kernel.org, linux-nvdimm@lists.01.org
Subject: [PATCH] x86/mce: Gate copy_mc_fragile() export by CONFIG_COPY_MC_TEST=y
Date: Wed, 07 Oct 2020 11:24:32 -0700	[thread overview]
Message-ID: <160209507277.2768223.9933672492157583642.stgit@dwillia2-desk3.amr.corp.intel.com> (raw)
In-Reply-To: <20201007111447.GA23257@zn.tnic>

It appears that modpost is not happy about exporting assembly symbols
that are not consumed in the same build. As Boris reports:

    WARNING: modpost: EXPORT symbol "copy_mc_fragile" [vmlinux] version generation failed, symbol will not be versioned.

The export is only consumed in the CONFIG_COPY_MC_TEST=y case, and even
then not in a way that modpost could see. CONFIG_COPY_MC_TEST uses a
module built in tools/testing/nvdimm/ to exercise the copy_mc_fragile()
corner cases.  Given the test already requires manually editing the
config entry for CONFIG_COPY_MC_TEST to make it "def_bool y" the
additional dependency to require is CONFIG_MODVERSIONS=n is not too
onerous.

Alternatively, COPY_MC_TEST and its related infrastructure could just be
ripped out because it has served its purpose. For now, just stop
exporting the symbol by default, and add the MODVERSIONS dependency to
the test.

Fixes: ec6347bb4339 ("x86, powerpc: Rename memcpy_mcsafe() to copy_mc_to_{user, kernel}()")
Reported-by: Borislav Petkov <bp@suse.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: x86@kernel.org
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Tony Luck <tony.luck@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 arch/x86/Kconfig.debug    |    1 +
 arch/x86/lib/copy_mc_64.S |    2 ++
 2 files changed, 3 insertions(+)

diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index 27b5e2bc6a01..6f0f5d8ac62e 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -63,6 +63,7 @@ config EARLY_PRINTK_USB_XDBC
 	  crashes or need a very simple printk logging facility.
 
 config COPY_MC_TEST
+	depends on !MODVERSIONS
 	def_bool n
 
 config EFI_PGT_DUMP
diff --git a/arch/x86/lib/copy_mc_64.S b/arch/x86/lib/copy_mc_64.S
index 892d8915f609..50fb05256751 100644
--- a/arch/x86/lib/copy_mc_64.S
+++ b/arch/x86/lib/copy_mc_64.S
@@ -88,7 +88,9 @@ SYM_FUNC_START(copy_mc_fragile)
 .L_done:
 	ret
 SYM_FUNC_END(copy_mc_fragile)
+#ifdef CONFIG_COPY_MC_TEST
 EXPORT_SYMBOL_GPL(copy_mc_fragile)
+#endif
 
 	.section .fixup, "ax"
 	/*


  parent reply	other threads:[~2020-10-07 20:00 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-18 20:30 [PATCH] x86/memcpy: Introduce memcpy_mcsafe_fast Andy Lutomirski
2020-04-18 20:30 ` Andy Lutomirski
2020-04-18 20:52 ` Linus Torvalds
2020-04-18 20:52   ` Linus Torvalds
2020-04-20  5:08   ` Dan Williams
2020-04-20  5:08     ` Dan Williams
2020-04-20 17:28     ` Linus Torvalds
2020-04-20 17:28       ` Linus Torvalds
2020-04-20 18:20       ` Dan Williams
2020-04-20 18:20         ` Dan Williams
2020-04-20 19:05         ` Linus Torvalds
2020-04-20 19:05           ` Linus Torvalds
2020-04-20 19:29           ` Dan Williams
2020-04-20 19:29             ` Dan Williams
2020-04-20 20:07             ` Linus Torvalds
2020-04-20 20:07               ` Linus Torvalds
2020-04-20 20:23               ` Luck, Tony
2020-04-20 20:23                 ` Luck, Tony
2020-04-20 20:27                 ` Linus Torvalds
2020-04-20 20:27                   ` Linus Torvalds
2020-04-20 20:45                   ` Luck, Tony
2020-04-20 20:45                     ` Luck, Tony
2020-04-20 20:56                     ` Linus Torvalds
2020-04-20 20:56                       ` Linus Torvalds
2020-04-20 20:24               ` Dan Williams
2020-04-20 20:24                 ` Dan Williams
2020-04-20 20:46                 ` Linus Torvalds
2020-04-20 20:46                   ` Linus Torvalds
2020-04-20 20:57                   ` Luck, Tony
2020-04-20 20:57                     ` Luck, Tony
2020-04-20 21:16                     ` Linus Torvalds
2020-04-20 21:16                       ` Linus Torvalds
2020-10-06  9:57       ` [tip: ras/core] x86, powerpc: Rename memcpy_mcsafe() to copy_mc_to_{user, kernel}() tip-bot2 for Dan Williams
2020-10-07 11:14         ` Borislav Petkov
2020-10-07 16:45           ` Borislav Petkov
2020-10-07 17:03             ` Borislav Petkov
2020-10-07 18:53               ` Dan Williams
2020-10-07 19:25                 ` Borislav Petkov
2020-10-08 16:59                   ` Dan Williams
2020-10-08 17:08                     ` Borislav Petkov
2020-10-07 17:51             ` Dan Williams
2020-10-07 18:24           ` Dan Williams [this message]
2020-10-07 18:24             ` [PATCH] x86/mce: Gate copy_mc_fragile() export by CONFIG_COPY_MC_TEST=y Dan Williams
2020-10-08  9:01           ` [tip: ras/core] x86/mce: Allow for copy_mc_fragile symbol checksum to be generated tip-bot2 for Borislav Petkov

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=160209507277.2768223.9933672492157583642.stgit@dwillia2-desk3.amr.corp.intel.com \
    --to=dan.j.williams@intel.com \
    --cc=bp@alien8.de \
    --cc=bp@suse.de \
    --cc=hpa@zytor.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nvdimm@lists.01.org \
    --cc=mingo@redhat.com \
    --cc=tglx@linutronix.de \
    --cc=tony.luck@intel.com \
    --cc=x86@kernel.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.