linux-kernel - Re: [PATCH] firmware_loader: Add debug message with checksum for FW file

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Y/i1z4ozr4P61FUh@kroah.com>
Date:   Fri, 24 Feb 2023 14:04:15 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Amadeusz Sławiński 
        <amadeuszx.slawinski@...ux.intel.com>
Cc:     Russ Weight <russell.h.weight@...el.com>,
        Luis Chamberlain <mcgrof@...nel.org>,
        linux-kernel@...r.kernel.org,
        Cezary Rojewski <cezary.rojewski@...el.com>,
        "Rafael J . Wysocki" <rafael.j.wysocki@...el.com>
Subject: Re: [PATCH] firmware_loader: Add debug message with checksum for FW
 file

On Fri, Feb 24, 2023 at 01:54:33PM +0100, Amadeusz Sławiński wrote:
> On 2/24/2023 1:46 PM, Greg Kroah-Hartman wrote:
> > On Fri, Feb 24, 2023 at 09:19:18PM +0100, Amadeusz Sławiński wrote:
> > > Enable dynamic-debug logging of firmware filenames and SHA256 checksums
> > > to clearly identify the firmware files that are loaded by the system.
> > > 
> > > Example output:
> > > [   34.944619] firmware_class:_request_firmware: i915 0000:00:02.0: Loaded FW: i915/kbl_dmc_ver1_04.bin, sha256: 2cde41c3e5ad181423bcc3e98ff9c49f743c88f18646af4d0b3c3a9664b831a1
> > > [   48.155884] firmware_class:_request_firmware: snd_soc_avs 0000:00:1f.3: Loaded FW: intel/avs/cnl/dsp_basefw.bin, sha256: 43f6ac1b066e9bd0423d914960fbbdccb391af27d2b1da1085eee3ea8df0f357
> > > [   49.579540] firmware_class:_request_firmware: snd_soc_avs 0000:00:1f.3: Loaded FW: intel/avs/rt274-tplg.bin, sha256: 4b3580da96dc3d2c443ba20c6728d8b665fceb3ed57223c3a57582bbad8e2413
> > > [   49.798196] firmware_class:_request_firmware: snd_soc_avs 0000:00:1f.3: Loaded FW: intel/avs/hda-8086280c-tplg.bin, sha256: 5653172579b2be1b51fd69f5cf46e2bac8d63f2a1327924311c13b2f1fe6e601
> > > [   49.859627] firmware_class:_request_firmware: snd_soc_avs 0000:00:1f.3: Loaded FW: intel/avs/dmic-tplg.bin, sha256: 00fb7fbdb74683333400d7e46925dae60db448b88638efcca0b30215db9df63f
> > > 
> > > Reviewed-by: Russ Weight <russell.h.weight@...el.com>
> > > Reviewed-by: Cezary Rojewski <cezary.rojewski@...el.com>
> > > Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@...ux.intel.com>
> > > ---
> > >   drivers/base/firmware_loader/main.c | 45 ++++++++++++++++++++++++++++-
> > >   1 file changed, 44 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/base/firmware_loader/main.c b/drivers/base/firmware_loader/main.c
> > > index 017c4cdb219e..a6e1fb10763d 100644
> > > --- a/drivers/base/firmware_loader/main.c
> > > +++ b/drivers/base/firmware_loader/main.c
> > > @@ -791,6 +791,47 @@ static void fw_abort_batch_reqs(struct firmware *fw)
> > >   	mutex_unlock(&fw_lock);
> > >   }
> > > +#if defined(DEBUG) || defined(CONFIG_DYNAMIC_DEBUG)
> > > +#include <crypto/hash.h>
> > > +#include <crypto/sha2.h>
> > > +#define SHA256_STRING_SIZE (SHA256_DIGEST_SIZE * 2)
> > > +static void fw_log_firmware_info(const struct firmware *fw, const char *name, struct device *device)
> > > +{
> > > +	char outbuf[SHA256_STRING_SIZE + 1];
> > > +	u8 sha256buf[SHA256_DIGEST_SIZE];
> > 
> > Nit, these are big, are you _SURE_ you can put them on the stack ok?
> > Why not dynamically allocate them?
> > 
> 
> Well, those arrays are not that big? First one is 65 bytes and other one 32.

And how far down in the stack are you when a driver is requesting
firmware to be loaded?  Usually pretty deep :(

> Although now that I looked again at the header, there is SHA256_BLOCK_SIZE
> define for string size, so I will change SHA256_STRING_SIZE to that instead.

You are already dynamically allocating memory, just do it for all of
these please so we don't have to go fix this in a few years when some
codepath from a driver is found to have blown up the stack for this
debug information.

thanks,

greg k-h