Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 15 May 2018 23:32:45 +0100
From:   "Maciej W. Rozycki" <macro@mips.com>
To:     Alexander Viro <viro@zeniv.linux.org.uk>,
        James Hogan <jhogan@kernel.org>
CC:     Ralf Baechle <ralf@linux-mips.org>,
        <linux-fsdevel@vger.kernel.org>, <linux-mips@linux-mips.org>,
        <linux-kernel@vger.kernel.org>, <stable@vger.kernel.org>
Subject: [PATCH 1/3] binfmt_elf: Respect error return from `regset->active'
In-Reply-To: <alpine.DEB.2.00.1804301557320.11756@tp.orcam.me.uk>
Message-ID: <alpine.DEB.2.00.1805102009380.10896@tp.orcam.me.uk>
References: <alpine.DEB.2.00.1804301557320.11756@tp.orcam.me.uk>
User-Agent: Alpine 2.00 (DEB 1167 2008-08-23)
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

The regset API documented in <linux/regset.h> defines -ENODEV as the 
result of the `->active' handler to be used where the feature requested 
is not available on the hardware found.  However code handling core file 
note generation in `fill_thread_core_info' interpretes any non-zero 
result from the `->active' handler as the regset requested being active. 
Consequently processing continues (and hopefully gracefully fails later 
on) rather than being abandoned right away for the regset requested.

Fix the problem then by making the code proceed only if a positive 
result is returned from the `->active' handler.

Cc: stable@vger.kernel.org # 2.6.25+
Fixes: 4206d3aa1978 ("elf core dump: notes user_regset")
Signed-off-by: Maciej W. Rozycki <macro@mips.com>
---
Hi,

 Overall we could also use the count returned by ->active to limit the 
size of data requested, i.e. something along the lines of:

	ssize_t size;

	if (regset->active)
		size = regset->active(t->task, regset);
	else
		size = regset_size(t->task, regset);
	if (size > 0) {
		...	

however that would be an optimisation that belongs to a separate change, 
which (due to the need to unload stuff I have in progress already) I am 
not going to make at this point.  Perhaps someone else would be willing 
to pick up the idea.

 Anyway, please apply.

  Maciej
---
 fs/binfmt_elf.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

linux-elf-core-regset-active.diff
Index: linux-jhogan-test/fs/binfmt_elf.c
===================================================================
--- linux-jhogan-test.orig/fs/binfmt_elf.c	2018-03-21 17:14:55.000000000 +0000
+++ linux-jhogan-test/fs/binfmt_elf.c	2018-05-09 23:25:50.742255000 +0100
@@ -1739,7 +1739,7 @@ static int fill_thread_core_info(struct 
 		const struct user_regset *regset = &view->regsets[i];
 		do_thread_regset_writeback(t->task, regset);
 		if (regset->core_note_type && regset->get &&
-		    (!regset->active || regset->active(t->task, regset))) {
+		    (!regset->active || regset->active(t->task, regset) > 0)) {
 			int ret;
 			size_t size = regset_size(t->task, regset);
 			void *data = kmalloc(size, GFP_KERNEL);