qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Alexander Bulekov <alxndr@bu.edu>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: "qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Subject: Re: [PATCH v4 15/20] fuzz: add fuzzer skeleton
Date: Tue, 12 Nov 2019 14:04:12 -0500	[thread overview]
Message-ID: <488e43a4-91f9-69a0-93a0-97553d57cbb7@bu.edu> (raw)
In-Reply-To: <20191107125540.GA365089@stefanha-x1.localdomain>

On 11/7/19 7:55 AM, Stefan Hajnoczi wrote:
> On Wed, Oct 30, 2019 at 02:50:00PM +0000, Oleinik, Alexander wrote:
>> diff --git a/tests/fuzz/fuzz.c b/tests/fuzz/fuzz.c
>> new file mode 100644
>> index 0000000000..0e38f81c48
>> --- /dev/null
>> +++ b/tests/fuzz/fuzz.c
>> @@ -0,0 +1,177 @@
>> +/*
>> + * fuzzing driver
>> + *
>> + * Copyright Red Hat Inc., 2019
>> + *
>> + * Authors:
>> + *  Alexander Bulekov   <alxndr@bu.edu>
> 
> Bulekov instead of Oleinik?
Yes I changed my last name and the approval from the court finally came 
through last week :)
I'll make sure its consistent across v5.

>> + *
>> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
>> + * See the COPYING file in the top-level directory.
>> + *
>> + */
>> +
>> +#include "qemu/osdep.h"
>> +
>> +#include <stdio.h>
>> +#include <stdlib.h>
> 
> stdio.h and stdlib.h are already included by qemu/osdep.h.
> 
>> +/* Executed for each fuzzing-input */
>> +int LLVMFuzzerTestOneInput(const unsigned char *Data, size_t Size)
>> +{
>> +    if (fuzz_target->fuzz) {
> 
> Will this ever be NULL?
I'll remove the check

>> +        fuzz_target->fuzz(fuzz_qts, Data, Size);
>> +    }
>> +    return 0;
>> +}
>> +
>> +/* Executed once, prior to fuzzing */
>> +int LLVMFuzzerInitialize(int *argc, char ***argv, char ***envp)
>> +{
>> +
>> +    char *target_name;
>> +
>> +    /* Initialize qgraph and modules */
>> +    qos_graph_init();
>> +    module_call_init(MODULE_INIT_FUZZ_TARGET);
>> +    module_call_init(MODULE_INIT_QOM);
>> +    module_call_init(MODULE_INIT_LIBQOS);
>> +
>> +    if (*argc <= 1) {
>> +        usage(**argv);
>> +    }
>> +
>> +    /* Identify the fuzz target */
>> +    target_name = (*argv)[1];
>> +    if (!strstr(target_name, "--fuzz-target=")) {
>> +        usage(**argv);
>> +    }
>> +
>> +    target_name += strlen("--fuzz-target=");
>> +
>> +    fuzz_target = fuzz_get_target(target_name);
>> +    if (!fuzz_target) {
>> +        usage(**argv);
>> +    }
>> +
>> +    fuzz_qts = qtest_setup();
>> +
>> +    if (!fuzz_target) {
> 
> This is dead code since fuzz_target was already checked above.  Please
> remove this if statement.
> 
>> +        fprintf(stderr, "Error: Fuzz fuzz_target name %s not found\n",
>> +                target_name);
>> +        usage(**argv);
>> +    }
>> +
>> +    if (fuzz_target->pre_vm_init) {
>> +        fuzz_target->pre_vm_init();
>> +    }
>> +
>> +    /* Run QEMU's softmmu main with the fuzz-target dependent arguments */
>> +    char *init_cmdline = fuzz_target->get_init_cmdline(fuzz_target);
> 
> Where is init_cmdline freed or should this be const char *?
> 
>> +    wordexp_t result;
>> +    wordexp(init_cmdline, &result, 0);
> 
> What is the purpose of word expansion here?
The fuzz target devs can specify arguments in a single string and not 
worry about calculating the argc and **argv - we take care of it for them.

>> +
>> +    qemu_init(result.we_wordc, result.we_wordv, NULL);
>> +
>> +    if (fuzz_target->pre_fuzz) {
>> +        fuzz_target->pre_fuzz(fuzz_qts);
>> +    }
>> +
>> +    return 0;
>> +}
>> diff --git a/tests/fuzz/fuzz.h b/tests/fuzz/fuzz.h
>> new file mode 100644
>> index 0000000000..b569b622d7
>> --- /dev/null
>> +++ b/tests/fuzz/fuzz.h
>> @@ -0,0 +1,66 @@
>> +/*
>> + * fuzzing driver
>> + *
>> + * Copyright Red Hat Inc., 2019
>> + *
>> + * Authors:
>> + *  Alexander Bulekov   <alxndr@bu.edu>
>> + *
>> + * This work is licensed under the terms of the GNU GPL, version 2 or later.
>> + * See the COPYING file in the top-level directory.
>> + *
>> + */
>> +
>> +#ifndef FUZZER_H_
>> +#define FUZZER_H_
>> +
>> +#include "qemu/osdep.h"
>> +#include "qemu/units.h"
>> +#include "qapi/error.h"
>> +#include "exec/memory.h"
>> +#include "tests/libqtest.h"
>> +
>> +
> 
> Some documentation would be nice:
> 
...
> Does the caller need to call g_free() on the returned string?  Please
> document this.
...
> s/to to/to/
...
> Please also mention that QEMU has been initialized at this point.
> 
...
> "makes a copy of *target" -> does this mean the argument type can be
> const FuzzTarget *target?
> 

Thanks - I made changes to address these.
-Alex

-- 
===
I recently changed my last name from Oleinik to Bulekov
===


  reply	other threads:[~2019-11-12 19:05 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-30 14:49 [PATCH v4 00/20] Add virtual device fuzzing support Oleinik, Alexander
2019-10-30 14:49 ` [PATCH v4 01/20] softmmu: split off vl.c:main() into main.c Oleinik, Alexander
2019-11-05 16:41   ` Darren Kenny
2019-11-12 16:46     ` Alexander Bulekov
2019-11-06 15:01   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 02/20] libqos: Rename i2c_send and i2c_recv Oleinik, Alexander
2019-11-06 15:17   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 03/20] fuzz: Add FUZZ_TARGET module type Oleinik, Alexander
2019-11-06 13:17   ` Darren Kenny
2019-11-06 15:18   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 04/20] qtest: add qtest_server_send abstraction Oleinik, Alexander
2019-11-06 13:29   ` Darren Kenny
2019-11-06 15:19   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 06/20] module: check module wasn't already initialized Oleinik, Alexander
2019-11-06 15:26   ` Stefan Hajnoczi
2019-11-06 17:40   ` Darren Kenny
2019-10-30 14:49 ` [PATCH v4 05/20] libqtest: Add a layer of abstraciton to send/recv Oleinik, Alexander
2019-11-06 16:22   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 07/20] qtest: add in-process incoming command handler Oleinik, Alexander
2019-11-06 16:33   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 08/20] tests: provide test variables to other targets Oleinik, Alexander
2019-11-07 14:32   ` Darren Kenny
2019-10-30 14:49 ` [PATCH v4 09/20] libqos: split qos-test and libqos makefile vars Oleinik, Alexander
2019-11-07 14:03   ` Darren Kenny
2019-10-30 14:49 ` [PATCH v4 10/20] libqos: move useful qos-test funcs to qos_external Oleinik, Alexander
2019-11-06 16:41   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 11/20] libqtest: make qtest_bufwrite send "atomic" Oleinik, Alexander
2019-11-06 16:44   ` Stefan Hajnoczi
2019-10-30 14:49 ` [PATCH v4 12/20] libqtest: add in-process qtest.c tx/rx handlers Oleinik, Alexander
2019-11-06 16:56   ` Stefan Hajnoczi
2019-11-12 17:38     ` Alexander Bulekov
2019-10-30 14:49 ` [PATCH v4 13/20] fuzz: add configure flag --enable-fuzzing Oleinik, Alexander
2019-11-06 16:57   ` Stefan Hajnoczi
2019-10-30 14:50 ` [PATCH v4 15/20] fuzz: add fuzzer skeleton Oleinik, Alexander
2019-11-07 12:55   ` Stefan Hajnoczi
2019-11-12 19:04     ` Alexander Bulekov [this message]
2019-10-30 14:50 ` [PATCH v4 14/20] fuzz: Add target/fuzz makefile rules Oleinik, Alexander
2019-11-07 14:31   ` Darren Kenny
2019-10-30 14:50 ` [PATCH v4 16/20] fuzz: add support for fork-based fuzzing Oleinik, Alexander
2019-11-07 13:17   ` Stefan Hajnoczi
2019-10-30 14:50 ` [PATCH v4 17/20] fuzz: add support for qos-assisted fuzz targets Oleinik, Alexander
2019-11-07 13:22   ` Stefan Hajnoczi
2019-10-30 14:50 ` [PATCH v4 18/20] fuzz: add i440fx " Oleinik, Alexander
2019-11-07 13:26   ` Stefan Hajnoczi
2019-10-30 14:50 ` [PATCH v4 19/20] fuzz: add virtio-net fuzz target Oleinik, Alexander
2019-11-07 13:36   ` Stefan Hajnoczi
2019-11-07 13:42   ` Jason Wang
2019-11-07 15:41     ` Stefan Hajnoczi
2019-10-30 14:50 ` [PATCH v4 20/20] fuzz: add documentation to docs/devel/ Oleinik, Alexander
2019-11-07 13:40   ` Stefan Hajnoczi
2019-11-07 15:02     ` Alexander Oleinik
2019-10-30 15:23 ` [PATCH v4 00/20] Add virtual device fuzzing support no-reply
2019-11-06 15:27   ` Stefan Hajnoczi
2019-11-05 13:57 ` Darren Kenny
2019-11-05 16:28   ` Alexander Oleinik
2019-11-05 16:47     ` Darren Kenny
2019-11-07 13:41 ` Stefan Hajnoczi

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=488e43a4-91f9-69a0-93a0-97553d57cbb7@bu.edu \
    --to=alxndr@bu.edu \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@gmail.com \
    /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 NNTP newsgroup(s).