Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Subject: Re: [PATCH 4/4] file: Fix file descriptor leak in copy_fd_bitmaps()
To:     Fedor Pchelkin <aissur0002@gmail.com>,
        Alexander Viro <viro@zeniv.linux.org.uk>
Cc:     linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        Linus Torvalds <torvalds@linux-foundation.org>
References: <20220326114009.1690-1-aissur0002@gmail.com>
From:   Alexey Khoroshilov <khoroshilov@ispras.ru>
Autocrypt: addr=khoroshilov@ispras.ru; prefer-encrypt=mutual; keydata=
 xsFNBFtq9eIBEACxmOIPDht+aZvO9DGi4TwnZ1WTDnyDVz3Nnh0rlQCK8IssaT6wE5a95VWo
 iwOWalcL9bJMHQvw60JwZKFjt9oH2bov3xzx/JRCISQB4a4U1J/scWvPtabbB3t+VAodF5KZ
 vZ2gu/Q/Wa5JZ9aBH0IvNpBAAThFg1rBXKh7wNqrhsQlMLg+zTSK6ZctddNl6RyaJvAmbaTS
 sSeyUKXiabxHn3BR9jclXfmPLfWuayinBvW4J3vS+bOhbLxeu3MO0dUqeX/Nl8EAhvzo0I2d
 A0vRu/Ze1wU3EQYT6M8z3i1b3pdLjr/i+MI8Rgijs+TFRAhxRw/+0vHGTg6Pn02t0XkycxQR
 mhH3v0kVTvMyM7YSI7yXvd0QPxb1RX9AGmvbJu7eylzcq9Jla+/T3pOuWsJkbvbvuFKKmmYY
 WnAOR7vu/VNVfiy4rM0bfO14cIuEG+yvogcPuMmQGYu6ZwS9IdgZIOAkO57M/6wR0jIyfxrG
 FV3ietPtVcqeDVrcShKyziRLJ+Xcsg9BLdnImAqVQomYr27pyNMRL5ILuT7uOuAQPDKBksK+
 l2Fws0d5iUifqnXSPuYxqgS4f8SQLS7ECxvCGVVbkEEng9vkkmyrF6wM86BZ9apPGDFbopiK
 7GRxQtSGszVv83abaVb8aDsAudJIp7lLaIuXLZAe1r+ycYpEtQARAQABzSpBbGV4ZXkgS2hv
 cm9zaGlsb3YgPGtob3Jvc2hpbG92QGlzcHJhcy5ydT7CwX0EEwEIACcFAltq9eICGwMFCRLM
 AwAFCwkIBwIGFQgJCgsCBBYCAwECHgECF4AACgkQ2B/JSzCwrEWLaA/+NFZfyhU0vJzFtYsk
 yaqx8nWZLrAoUK7VcobH0lJH6lfGbarO5JpENaIiTP12YZ4xO+j3GGJtLy2gvnpypGnxmiAl
 RqPt7WeAIj6oqPrUs2QF7i4SOiPtku/NrysI1zHzlA8yqUduBtam5rdQeLRNCJiEED1fU8sp
 +DgJBN/OHEDyAag2hu1KFKWuPfQ+QGpXYZb+1NW/hKwvvwCNVyypELAfFnkketFXjIMwHnL8
 ZPqJZlkvkpxuRXOaXPL9NFhZnC/WS+NJ81L3pr+w6eo3xTPYZvRW8glvqlEDgHqr3uMGIaes
 nwfRXLHp+TC1ht6efCXzdPyMZ1E7HXQN9foKisI1V5iQFhN+CT3dbsguQI4e10F5ql0TZUJY
 SMzvY0eObs6TWRdD/Ha7Y5rLmZ54R9sxumpZNcJzktfgm9f0XfeqVEJUn/40MRDD+l2W12Db
 Jkko+sbtAEw+f+/j3uz8xOE+Uv4kwFC5a6JKgdX88oigHnpAs3FvffP594Loi3ibFrQUW5wH
 bXh5Ni+l1GKEQ0PHMk+KQQT9L2r9s7C0Nh8XzwdpOshZWsrNSZqcG+01wrmUhyX2uSaoZ07I
 /+KZURlMSqI71X6lkMWlB3SyThvYhHgnR0EGGTerwM1MaVjHN+Z6lPmsKNxG8lzCeWeZ6peA
 c5oUHV4WQ8Ux9BM8saLOwU0EW2r14gEQAMz+5u+X7j1/dT4WLVRQaE1Shnd2dKBn2E7fgo/N
 4JIY6wHD/DJoWYQpCJjjvBYSonvQsHicvDW8lPh2EXgZ9Fi8AHKT2mVPitVy+uhfWa/0FtsC
 e3hPfrjTcN7BUcXlIjmptxIoDbvQrNfIWUGdWiyDj4EDfABW/kagXqaBwF2HdcDaNDGggD1c
 DglA0APjezIyTGnGMKsi5QSSlOLm8OZEJMj5t+JL6QXrruijNb5Asmz5mpRQrak7DpGOskjK
 fClm/0oy2zDvWuoXJa+dm3YFr43V+c5EIMA4LpGk63Eg+5NltQ/gj0ycgD5o6reCbjLz4R9D
 JzBezK/KOQuNG5qKUTMbOHWaApZnZ6BDdOVflkV1V+LMo5GvIzkATNLm/7Jj6DmYmXbKoSAY
 BKZiJWqzNsL1AJtmJA1y5zbWX/W4CpNs8qYMYG8eTNOqunzopEhX7T0cOswcTGArZYygiwDW
 BuIS83QRc7udMlQg79qyMA5WqS9g9g/iodlssR9weIVoZSjfjhm5NJ3FmaKnb56h6DSvFgsH
 xCa4s1DGnZGSAtedj8E3ACOsEfu4J/WqXEmvMYNBdGos2YAc+g0hjuOB10BSD98d38xP1vPc
 qNrztIF+TODAl1dNwU4rCSdGQymsrMVFuXnHMH4G+dHvMAwWauzDbnILHAGFyJtfxVefABEB
 AAHCwWUEGAEIAA8FAltq9eICGwwFCRLMAwAACgkQ2B/JSzCwrEU3Rg//eFWHXqTQ5CKw4KrX
 kTFxdXnYKJ5zZB0EzqU6m/FAV7snmygFLbOXYlcMW2Fh306ivj9NKJrlOaPbUzzyDf8dtDAg
 nSbH156oNJ9NHkz0mrxFMpJA2E5AUemOFx57PUYt93pR2B7bF2zGua4gMC+vorDQZjX9kvrL
 Kbenh3boFOe1tUaiRRvEltVFLOg+b+CMkKVbLIQe/HkyKJH5MFiHAF7QxnPHaxyO7QbWaUmF
 6BHVujxAGvNgkrYJb6dpiNNZSFNRodaSToU5oM+z1dCrNNtN3u4R7AYr6DDIDxoSzR4k0ZaG
 uSeqh4xxQCD7vLT3JdZDyhYUJgy9mvSXdkXGdBIhVmeLch2gaWNf5UOutVJwdPbIaUDRjVoV
 Iw6qjKq+mnK3ttuxW5Aeg9Y1OuKEvCVu+U/iEEJxx1JRmVAYq848YqtVPY9DkZdBT4E9dHqO
 n8lr+XPVyMN6SBXkaR5tB6zSkSDrIw+9uv1LN7QIri43fLqhM950ltlveROEdLL1bI30lYO5
 J07KmxgOjrvY8X9WOC3O0k/nFpBbbsM4zUrmF6F5wIYO99xafQOlfpUnVtbo3GnBR2LIcPYj
 SyY3dW28JXo2cftxIOr1edJ+fhcRqYRrPzJrQBZcE2GZjRO8tz6IOMAsc+WMtVfj5grgVHCu
 kK2E04Fb+Zk1eJvHYRc=
Message-ID: <c7fcaccf-7ac0-fae8-3f41-d6552b689a70@ispras.ru>
Date:   Sat, 26 Mar 2022 17:17:32 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <20220326114009.1690-1-aissur0002@gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

Looks like bfp has a set of macro suitable for such cases:

#define BITS_PER_BYTE_MASKED(bits) ((bits) & BITS_PER_BYTE_MASK)
#define BITS_ROUNDDOWN_BYTES(bits) ((bits) >> 3)
#define BITS_ROUNDUP_BYTES(bits) \
	(BITS_ROUNDDOWN_BYTES(bits) + !!BITS_PER_BYTE_MASKED(bits))

May be it makes sense to move them to a generic header and to use here?

--
Alexey Khoroshilov


On 26.03.2022 14:40, Fedor Pchelkin wrote:
> If count argument in copy_fd_bitmaps() is not a multiple of
> BITS_PER_BYTE, then one byte is lost and is not used in further
> manipulations with cpy value in memcpy() and memset()
> causing a leak. The leak was introduced with close_range() call
> using CLOSE_RANGE_UNSHARE flag.
> 
> The patch suggests implementing an indicator (named add_byte)
> of count being multiple of BITS_PER_BYTE and adding it to the
> cpy value.
> 
> Found by Syzkaller (https://github.com/google/syzkaller).
> 
> Signed-off-by: Fedor Pchelkin <aissur0002@gmail.com>
> Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
> ---
>  fs/file.c | 2 --
>  1 file changed, 2 deletions(-)
> 
> diff --git a/fs/file.c b/fs/file.c
> index 3ef1479df203..3c64a6423604 100644
> --- a/fs/file.c
> +++ b/fs/file.c
> @@ -56,10 +56,8 @@ static void copy_fd_bitmaps(struct fdtable *nfdt, struct fdtable *ofdt,
>  {
>  	unsigned int cpy, set;
>  	unsigned int add_byte = 0;
> -	
>  	if (count % BITS_PER_BYTE != 0)
>  		add_byte = 1;
> -	
>  	cpy = count / BITS_PER_BYTE + add_byte;
>  	set = (nfdt->max_fds - count) / BITS_PER_BYTE;
>  	memcpy(nfdt->open_fds, ofdt->open_fds, cpy);
>