Re: [PATCH AUTOSEL 5.2 13/85] kernel/module.c: Only return -EEXIST for modules that have finished loading

From: Prarit Bhargava <prarit@redhat.com>
To: Sasha Levin <sashal@kernel.org>,
	linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Barret Rhoden <brho@google.com>,
	David Arcari <darcari@redhat.com>, Jessica Yu <jeyu@kernel.org>,
	Heiko Carstens <heiko.carstens@de.ibm.com>
Subject: Re: [PATCH AUTOSEL 5.2 13/85] kernel/module.c: Only return -EEXIST for modules that have finished loading
Date: Fri, 26 Jul 2019 09:46:46 -0400	[thread overview]
Message-ID: <cb703dc8-03a3-a877-2e5f-72a349e0f2d7@redhat.com> (raw)
In-Reply-To: <20190726133936.11177-13-sashal@kernel.org>

On 7/26/19 9:38 AM, Sasha Levin wrote:
> From: Prarit Bhargava <prarit@redhat.com>
> 
> [ Upstream commit 6e6de3dee51a439f76eb73c22ae2ffd2c9384712 ]
> 
> Microsoft HyperV disables the X86_FEATURE_SMCA bit on AMD systems, and
> linux guests boot with repeated errors:
> 

Hey Sasha, I'd prefer to leave this out of stable branches for now.  Linus is a
bit nervous about it and I like to get see more soak time before the patch is
backported to stable.

https://lkml.org/lkml/2019/7/18/680

FWIW we've been running this in RHEL for some time now without any issues.

P.

> amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err -2)
> amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err -2)
> amd64_edac_mod: Unknown symbol amd_report_gart_errors (err -2)
> amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err -2)
> amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err -2)
> amd64_edac_mod: Unknown symbol amd_report_gart_errors (err -2)
> 
> The warnings occur because the module code erroneously returns -EEXIST
> for modules that have failed to load and are in the process of being
> removed from the module list.
> 
> module amd64_edac_mod has a dependency on module edac_mce_amd.  Using
> modules.dep, systemd will load edac_mce_amd for every request of
> amd64_edac_mod.  When the edac_mce_amd module loads, the module has
> state MODULE_STATE_UNFORMED and once the module load fails and the state
> becomes MODULE_STATE_GOING.  Another request for edac_mce_amd module
> executes and add_unformed_module() will erroneously return -EEXIST even
> though the previous instance of edac_mce_amd has MODULE_STATE_GOING.
> Upon receiving -EEXIST, systemd attempts to load amd64_edac_mod, which
> fails because of unknown symbols from edac_mce_amd.
> 
> add_unformed_module() must wait to return for any case other than
> MODULE_STATE_LIVE to prevent a race between multiple loads of
> dependent modules.
> 
> Signed-off-by: Prarit Bhargava <prarit@redhat.com>
> Signed-off-by: Barret Rhoden <brho@google.com>
> Cc: David Arcari <darcari@redhat.com>
> Cc: Jessica Yu <jeyu@kernel.org>
> Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
> Signed-off-by: Jessica Yu <jeyu@kernel.org>
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
>  kernel/module.c | 6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/kernel/module.c b/kernel/module.c
> index 80c7c09584cf..8431c3d47c97 100644
> --- a/kernel/module.c
> +++ b/kernel/module.c
> @@ -3385,8 +3385,7 @@ static bool finished_loading(const char *name)
>  	sched_annotate_sleep();
>  	mutex_lock(&module_mutex);
>  	mod = find_module_all(name, strlen(name), true);
> -	ret = !mod || mod->state == MODULE_STATE_LIVE
> -		|| mod->state == MODULE_STATE_GOING;
> +	ret = !mod || mod->state == MODULE_STATE_LIVE;
>  	mutex_unlock(&module_mutex);
>  
>  	return ret;
> @@ -3576,8 +3575,7 @@ static int add_unformed_module(struct module *mod)
>  	mutex_lock(&module_mutex);
>  	old = find_module_all(mod->name, strlen(mod->name), true);
>  	if (old != NULL) {
> -		if (old->state == MODULE_STATE_COMING
> -		    || old->state == MODULE_STATE_UNFORMED) {
> +		if (old->state != MODULE_STATE_LIVE) {
>  			/* Wait in case it fails to load. */
>  			mutex_unlock(&module_mutex);
>  			err = wait_event_interruptible(module_wq,
> 