* [PATCH] Input: cyttsp4 - Fix error on calculating memory size passed to krealloc. @ 2017-10-31 19:19 Vince Kim [not found] ` <CAJRpnq5CtGDHjSpa_mBqsJmZxvdrkAwPYwFk1tM20h4vvEbb4A@mail.gmail.com> 0 siblings, 1 reply; 4+ messages in thread From: Vince Kim @ 2017-10-31 19:19 UTC (permalink / raw) Cc: Vince Kim, Ferruh Yigit, Dmitry Torokhov, linux-input, linux-kernel There are several places to perform subtraction to calculate buffer size such as: si->si_ofs.cydata_size = si->si_ofs.test_ofs - si->si_ofs.cydata_ofs; ... p = krealloc(si->si_ptrs.cydata, si->si_ofs.cydata_size, GFP_KERNEL); Actually, data types of above variables during subtraction are size_t, so it is unsigned. That means if second operand(si->si_ofs.cydata_ofs) is greater than the first operand(si->si_ofs.test_ofs), then resulting si->si_ofs.cydata_size could result in an unsigned integer wrap which is not desiarable. The properway to correct this problem is to perform a test of both operands to avoid having unsigned wrap. --- drivers/input/touchscreen/cyttsp4_core.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/drivers/input/touchscreen/cyttsp4_core.c b/drivers/input/touchscreen/cyttsp4_core.c index beaf61c..eecc7f1 100644 --- a/drivers/input/touchscreen/cyttsp4_core.c +++ b/drivers/input/touchscreen/cyttsp4_core.c @@ -201,6 +201,11 @@ static int cyttsp4_si_get_cydata(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.test_ofs <= si->si_ofs.cydata_ofs) + dev_err(cd->dev, "%s: invalid offset test_ofs:%zd, cydata_ofs:%zd \n", __func__, si->si_ofs.test_ofs, si->si_ofs.cydata_ofs); + return -EINVAL; + } + si->si_ofs.cydata_size = si->si_ofs.test_ofs - si->si_ofs.cydata_ofs; dev_dbg(cd->dev, "%s: cydata size: %zd\n", __func__, si->si_ofs.cydata_size); @@ -270,6 +275,11 @@ static int cyttsp4_si_get_test_data(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.pcfg_ofs <= si->si_ofs.test_ofs) + dev_err(cd->dev, "%s: invalid offset pcfg_ofs:%zd, test_ofs:%zd \n", __func__, si->si_ofs.pcfg_ofs, si->si_ofs.test_ofs); + return -EINVAL; + } + si->si_ofs.test_size = si->si_ofs.pcfg_ofs - si->si_ofs.test_ofs; p = krealloc(si->si_ptrs.test, si->si_ofs.test_size, GFP_KERNEL); @@ -321,6 +331,11 @@ static int cyttsp4_si_get_pcfg_data(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.opcfg_ofs <= si->si_ofs.pcfg_ofs) + dev_err(cd->dev, "%s: invalid offset opcfg_ofs:%zd, pcfg_ofs:%zd \n", __func__, si->si_ofs.opcfg_ofs, si->si_ofs.pcfg_ofs); + return -EINVAL; + } + si->si_ofs.pcfg_size = si->si_ofs.opcfg_ofs - si->si_ofs.pcfg_ofs; p = krealloc(si->si_ptrs.pcfg, si->si_ofs.pcfg_size, GFP_KERNEL); @@ -367,6 +382,11 @@ static int cyttsp4_si_get_opcfg_data(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.ddata_ofs <= si->si_ofs.opcfg_ofs) + dev_err(cd->dev, "%s: invalid offset ddata_ofs:%zd, opcfg_ofs:%zd \n", __func__, si->si_ofs.ddata_ofs, si->si_ofs.opcfg_ofs); + return -EINVAL; + } + si->si_ofs.opcfg_size = si->si_ofs.ddata_ofs - si->si_ofs.opcfg_ofs; p = krealloc(si->si_ptrs.opcfg, si->si_ofs.opcfg_size, GFP_KERNEL); -- 2.7.4 ^ permalink raw reply related [flat|nested] 4+ messages in thread
[parent not found: <CAJRpnq5CtGDHjSpa_mBqsJmZxvdrkAwPYwFk1tM20h4vvEbb4A@mail.gmail.com>]
* Re: [PATCH] Input: cyttsp4 - Fix error on calculating memory size passed to krealloc. [not found] ` <CAJRpnq5CtGDHjSpa_mBqsJmZxvdrkAwPYwFk1tM20h4vvEbb4A@mail.gmail.com> @ 2017-11-02 23:54 ` Dmitry Torokhov [not found] ` <CAJRpnq5CdTk0UHhyN8D-n4U3cutA11KDPYXsRQATc+BHmpzNqQ@mail.gmail.com> 0 siblings, 1 reply; 4+ messages in thread From: Dmitry Torokhov @ 2017-11-02 23:54 UTC (permalink / raw) To: Vince Kim; +Cc: Ferruh Yigit, linux-input, linux-kernel Hi Vince, On Wed, Nov 01, 2017 at 09:13:58PM -0700, Vince Kim wrote: > It looks patch send to the maintainer rejected. > I can take the patch, but I need your signed-off-by please. > Can anyone take a look? > > > > *==============================* > > *Delivery has failed to these recipients or groups:* > > Ferruh Yigit (fery@cypress.com) <fery@cypress.com> > Your message wasn't delivered due to a permission or security issue. It may > have been rejected by a moderator, the address may only accept e-mail from > certain senders, or another restriction may be preventing delivery. > *==============================* > > On Tue, Oct 31, 2017 at 12:19 PM, Vince Kim <vince.k.kim@gmail.com> wrote: > > > There are several places to perform subtraction to calculate buffer > > size such as: > > > > si->si_ofs.cydata_size = si->si_ofs.test_ofs - si->si_ofs.cydata_ofs; > > ... > > p = krealloc(si->si_ptrs.cydata, si->si_ofs.cydata_size, GFP_KERNEL); > > > > Actually, data types of above variables during subtraction are size_t, so > > it is > > unsigned. That means if second operand(si->si_ofs.cydata_ofs) is > > greater than the first operand(si->si_ofs.test_ofs), then resulting > > si->si_ofs.cydata_size could result in an unsigned integer wrap which is > > not desiarable. > > > > The properway to correct this problem is to perform a test of both > > operands to avoid having unsigned wrap. > > --- > > drivers/input/touchscreen/cyttsp4_core.c | 20 ++++++++++++++++++++ > > 1 file changed, 20 insertions(+) > > > > diff --git a/drivers/input/touchscreen/cyttsp4_core.c > > b/drivers/input/touchscreen/cyttsp4_core.c > > index beaf61c..eecc7f1 100644 > > --- a/drivers/input/touchscreen/cyttsp4_core.c > > +++ b/drivers/input/touchscreen/cyttsp4_core.c > > @@ -201,6 +201,11 @@ static int cyttsp4_si_get_cydata(struct cyttsp4 *cd) > > void *p; > > int rc; > > > > + if (si->si_ofs.test_ofs <= si->si_ofs.cydata_ofs) > > + dev_err(cd->dev, "%s: invalid offset test_ofs:%zd, > > cydata_ofs:%zd \n", __func__, si->si_ofs.test_ofs, si->si_ofs.cydata_ofs); > > + return -EINVAL; > > + } > > + > > si->si_ofs.cydata_size = si->si_ofs.test_ofs - > > si->si_ofs.cydata_ofs; > > dev_dbg(cd->dev, "%s: cydata size: %zd\n", __func__, > > si->si_ofs.cydata_size); > > @@ -270,6 +275,11 @@ static int cyttsp4_si_get_test_data(struct cyttsp4 > > *cd) > > void *p; > > int rc; > > > > + if (si->si_ofs.pcfg_ofs <= si->si_ofs.test_ofs) > > + dev_err(cd->dev, "%s: invalid offset pcfg_ofs:%zd, > > test_ofs:%zd \n", __func__, si->si_ofs.pcfg_ofs, si->si_ofs.test_ofs); > > + return -EINVAL; > > + } > > + > > si->si_ofs.test_size = si->si_ofs.pcfg_ofs - si->si_ofs.test_ofs; > > > > p = krealloc(si->si_ptrs.test, si->si_ofs.test_size, GFP_KERNEL); > > @@ -321,6 +331,11 @@ static int cyttsp4_si_get_pcfg_data(struct cyttsp4 > > *cd) > > void *p; > > int rc; > > > > + if (si->si_ofs.opcfg_ofs <= si->si_ofs.pcfg_ofs) > > + dev_err(cd->dev, "%s: invalid offset opcfg_ofs:%zd, > > pcfg_ofs:%zd \n", __func__, si->si_ofs.opcfg_ofs, si->si_ofs.pcfg_ofs); > > + return -EINVAL; > > + } > > + > > si->si_ofs.pcfg_size = si->si_ofs.opcfg_ofs - si->si_ofs.pcfg_ofs; > > > > p = krealloc(si->si_ptrs.pcfg, si->si_ofs.pcfg_size, GFP_KERNEL); > > @@ -367,6 +382,11 @@ static int cyttsp4_si_get_opcfg_data(struct cyttsp4 > > *cd) > > void *p; > > int rc; > > > > + if (si->si_ofs.ddata_ofs <= si->si_ofs.opcfg_ofs) > > + dev_err(cd->dev, "%s: invalid offset ddata_ofs:%zd, > > opcfg_ofs:%zd \n", __func__, si->si_ofs.ddata_ofs, si->si_ofs.opcfg_ofs); > > + return -EINVAL; > > + } > > + > > si->si_ofs.opcfg_size = si->si_ofs.ddata_ofs - > > si->si_ofs.opcfg_ofs; > > > > p = krealloc(si->si_ptrs.opcfg, si->si_ofs.opcfg_size, GFP_KERNEL); > > -- > > 2.7.4 > > > > -- Dmitry ^ permalink raw reply [flat|nested] 4+ messages in thread
[parent not found: <CAJRpnq5CdTk0UHhyN8D-n4U3cutA11KDPYXsRQATc+BHmpzNqQ@mail.gmail.com>]
* Re: [PATCH] Input: cyttsp4 - Fix error on calculating memory size passed to krealloc. [not found] ` <CAJRpnq5CdTk0UHhyN8D-n4U3cutA11KDPYXsRQATc+BHmpzNqQ@mail.gmail.com> @ 2017-11-03 19:09 ` Dmitry Torokhov 2017-11-03 21:34 ` [PATCH v2] " Vince Kim 0 siblings, 1 reply; 4+ messages in thread From: Dmitry Torokhov @ 2017-11-03 19:09 UTC (permalink / raw) To: Vince Kim; +Cc: Ferruh Yigit, linux-input, linux-kernel On Thu, Nov 02, 2017 at 08:49:35PM -0700, Vince Kim wrote: > Hello Dmitry, > > I attached patch with signed-off-by. *sigh* did you try compile that? + if (si->si_ofs.test_ofs <= si->si_ofs.cydata_ofs) + dev_err(cd->dev, "%s: invalid offset test_ofs:%zd, cydata_ofs:%zd \n", __func__, si->si_ofs.test_ofs, si->si_ofs.cydata_ofs); + return -EINVAL; + } + There is an opening curly brace missing. Same goes for the other 3 chunks of the patch. I fixed it up. > If you need a patch inline, please let me know. The inline patch formatted according to Documentation/SubmittingPatches is preferred, but I'd settle on something that compiles. Thanks. > > Thanks, > Vince > > > > > On Thu, Nov 2, 2017 at 4:54 PM, Dmitry Torokhov <dmitry.torokhov@gmail.com> > wrote: > > > Hi Vince, > > > > On Wed, Nov 01, 2017 at 09:13:58PM -0700, Vince Kim wrote: > > > It looks patch send to the maintainer rejected. > > > > > > > I can take the patch, but I need your signed-off-by please. > > > > > Can anyone take a look? > > > > > > > > > > > > *==============================* > > > > > > *Delivery has failed to these recipients or groups:* > > > > > > Ferruh Yigit (fery@cypress.com) <fery@cypress.com> > > > Your message wasn't delivered due to a permission or security issue. It > > may > > > have been rejected by a moderator, the address may only accept e-mail > > from > > > certain senders, or another restriction may be preventing delivery. > > > *==============================* > > > > > > On Tue, Oct 31, 2017 at 12:19 PM, Vince Kim <vince.k.kim@gmail.com> > > wrote: > > > > > > > There are several places to perform subtraction to calculate buffer > > > > size such as: > > > > > > > > si->si_ofs.cydata_size = si->si_ofs.test_ofs - si->si_ofs.cydata_ofs; > > > > ... > > > > p = krealloc(si->si_ptrs.cydata, si->si_ofs.cydata_size, GFP_KERNEL); > > > > > > > > Actually, data types of above variables during subtraction are size_t, > > so > > > > it is > > > > unsigned. That means if second operand(si->si_ofs.cydata_ofs) is > > > > greater than the first operand(si->si_ofs.test_ofs), then resulting > > > > si->si_ofs.cydata_size could result in an unsigned integer wrap which > > is > > > > not desiarable. > > > > > > > > The properway to correct this problem is to perform a test of both > > > > operands to avoid having unsigned wrap. > > > > --- > > > > drivers/input/touchscreen/cyttsp4_core.c | 20 ++++++++++++++++++++ > > > > 1 file changed, 20 insertions(+) > > > > > > > > diff --git a/drivers/input/touchscreen/cyttsp4_core.c > > > > b/drivers/input/touchscreen/cyttsp4_core.c > > > > index beaf61c..eecc7f1 100644 > > > > --- a/drivers/input/touchscreen/cyttsp4_core.c > > > > +++ b/drivers/input/touchscreen/cyttsp4_core.c > > > > @@ -201,6 +201,11 @@ static int cyttsp4_si_get_cydata(struct cyttsp4 > > *cd) > > > > void *p; > > > > int rc; > > > > > > > > + if (si->si_ofs.test_ofs <= si->si_ofs.cydata_ofs) > > > > + dev_err(cd->dev, "%s: invalid offset test_ofs:%zd, > > > > cydata_ofs:%zd \n", __func__, si->si_ofs.test_ofs, > > si->si_ofs.cydata_ofs); > > > > + return -EINVAL; > > > > + } > > > > + > > > > si->si_ofs.cydata_size = si->si_ofs.test_ofs - > > > > si->si_ofs.cydata_ofs; > > > > dev_dbg(cd->dev, "%s: cydata size: %zd\n", __func__, > > > > si->si_ofs.cydata_size); > > > > @@ -270,6 +275,11 @@ static int cyttsp4_si_get_test_data(struct > > cyttsp4 > > > > *cd) > > > > void *p; > > > > int rc; > > > > > > > > + if (si->si_ofs.pcfg_ofs <= si->si_ofs.test_ofs) > > > > + dev_err(cd->dev, "%s: invalid offset pcfg_ofs:%zd, > > > > test_ofs:%zd \n", __func__, si->si_ofs.pcfg_ofs, si->si_ofs.test_ofs); > > > > + return -EINVAL; > > > > + } > > > > + > > > > si->si_ofs.test_size = si->si_ofs.pcfg_ofs - > > si->si_ofs.test_ofs; > > > > > > > > p = krealloc(si->si_ptrs.test, si->si_ofs.test_size, > > GFP_KERNEL); > > > > @@ -321,6 +331,11 @@ static int cyttsp4_si_get_pcfg_data(struct > > cyttsp4 > > > > *cd) > > > > void *p; > > > > int rc; > > > > > > > > + if (si->si_ofs.opcfg_ofs <= si->si_ofs.pcfg_ofs) > > > > + dev_err(cd->dev, "%s: invalid offset opcfg_ofs:%zd, > > > > pcfg_ofs:%zd \n", __func__, si->si_ofs.opcfg_ofs, si->si_ofs.pcfg_ofs); > > > > + return -EINVAL; > > > > + } > > > > + > > > > si->si_ofs.pcfg_size = si->si_ofs.opcfg_ofs - > > si->si_ofs.pcfg_ofs; > > > > > > > > p = krealloc(si->si_ptrs.pcfg, si->si_ofs.pcfg_size, > > GFP_KERNEL); > > > > @@ -367,6 +382,11 @@ static int cyttsp4_si_get_opcfg_data(struct > > cyttsp4 > > > > *cd) > > > > void *p; > > > > int rc; > > > > > > > > + if (si->si_ofs.ddata_ofs <= si->si_ofs.opcfg_ofs) > > > > + dev_err(cd->dev, "%s: invalid offset ddata_ofs:%zd, > > > > opcfg_ofs:%zd \n", __func__, si->si_ofs.ddata_ofs, > > si->si_ofs.opcfg_ofs); > > > > + return -EINVAL; > > > > + } > > > > + > > > > si->si_ofs.opcfg_size = si->si_ofs.ddata_ofs - > > > > si->si_ofs.opcfg_ofs; > > > > > > > > p = krealloc(si->si_ptrs.opcfg, si->si_ofs.opcfg_size, > > GFP_KERNEL); > > > > -- > > > > 2.7.4 > > > > > > > > > > > > -- > > Dmitry > > -- Dmitry ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH v2] Input: cyttsp4 - Fix error on calculating memory size passed to krealloc. 2017-11-03 19:09 ` Dmitry Torokhov @ 2017-11-03 21:34 ` Vince Kim 0 siblings, 0 replies; 4+ messages in thread From: Vince Kim @ 2017-11-03 21:34 UTC (permalink / raw) Cc: Vince Kim, Ferruh Yigit, Dmitry Torokhov, linux-input, linux-kernel There are several places to perform subtraction to calculate buffer size such as: si->si_ofs.cydata_size = si->si_ofs.test_ofs - si->si_ofs.cydata_ofs; ... p = krealloc(si->si_ptrs.cydata, si->si_ofs.cydata_size, GFP_KERNEL); Actually, data types of above variables during subtraction are size_t, so it is unsigned. That means if second operand(si->si_ofs.cydata_ofs) is greater than the first operand(si->si_ofs.test_ofs), then resulting si->si_ofs.cydata_size could result in an unsigned integer wrap which is not desiarable. The properway to correct this problem is to perform a test of both operands to avoid having unsigned wrap. Signed-off-by: Vince Kim <vince.k.kim@gmail.com> --- Cahnges in v2: - added missing opening curly brace at if statement drivers/input/touchscreen/cyttsp4_core.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/drivers/input/touchscreen/cyttsp4_core.c b/drivers/input/touchscreen/cyttsp4_core.c index beaf61c..e4a3743 100644 --- a/drivers/input/touchscreen/cyttsp4_core.c +++ b/drivers/input/touchscreen/cyttsp4_core.c @@ -201,6 +201,11 @@ static int cyttsp4_si_get_cydata(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.test_ofs <= si->si_ofs.cydata_ofs) { + dev_err(cd->dev, "%s: invalid offset test_ofs:%zd, cydata_ofs:%zd \n", __func__, si->si_ofs.test_ofs, si->si_ofs.cydata_ofs); + return -EINVAL; + } + si->si_ofs.cydata_size = si->si_ofs.test_ofs - si->si_ofs.cydata_ofs; dev_dbg(cd->dev, "%s: cydata size: %zd\n", __func__, si->si_ofs.cydata_size); @@ -270,6 +275,11 @@ static int cyttsp4_si_get_test_data(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.pcfg_ofs <= si->si_ofs.test_ofs) { + dev_err(cd->dev, "%s: invalid offset pcfg_ofs:%zd, test_ofs:%zd \n", __func__, si->si_ofs.pcfg_ofs, si->si_ofs.test_ofs); + return -EINVAL; + } + si->si_ofs.test_size = si->si_ofs.pcfg_ofs - si->si_ofs.test_ofs; p = krealloc(si->si_ptrs.test, si->si_ofs.test_size, GFP_KERNEL); @@ -321,6 +331,11 @@ static int cyttsp4_si_get_pcfg_data(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.opcfg_ofs <= si->si_ofs.pcfg_ofs) { + dev_err(cd->dev, "%s: invalid offset opcfg_ofs:%zd, pcfg_ofs:%zd \n", __func__, si->si_ofs.opcfg_ofs, si->si_ofs.pcfg_ofs); + return -EINVAL; + } + si->si_ofs.pcfg_size = si->si_ofs.opcfg_ofs - si->si_ofs.pcfg_ofs; p = krealloc(si->si_ptrs.pcfg, si->si_ofs.pcfg_size, GFP_KERNEL); @@ -367,6 +382,11 @@ static int cyttsp4_si_get_opcfg_data(struct cyttsp4 *cd) void *p; int rc; + if (si->si_ofs.ddata_ofs <= si->si_ofs.opcfg_ofs) { + dev_err(cd->dev, "%s: invalid offset ddata_ofs:%zd, opcfg_ofs:%zd \n", __func__, si->si_ofs.ddata_ofs, si->si_ofs.opcfg_ofs); + return -EINVAL; + } + si->si_ofs.opcfg_size = si->si_ofs.ddata_ofs - si->si_ofs.opcfg_ofs; p = krealloc(si->si_ptrs.opcfg, si->si_ofs.opcfg_size, GFP_KERNEL); -- 2.7.4 ^ permalink raw reply related [flat|nested] 4+ messages in thread
end of thread, other threads:[~2017-11-03 21:34 UTC | newest] Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2017-10-31 19:19 [PATCH] Input: cyttsp4 - Fix error on calculating memory size passed to krealloc Vince Kim [not found] ` <CAJRpnq5CtGDHjSpa_mBqsJmZxvdrkAwPYwFk1tM20h4vvEbb4A@mail.gmail.com> 2017-11-02 23:54 ` Dmitry Torokhov [not found] ` <CAJRpnq5CdTk0UHhyN8D-n4U3cutA11KDPYXsRQATc+BHmpzNqQ@mail.gmail.com> 2017-11-03 19:09 ` Dmitry Torokhov 2017-11-03 21:34 ` [PATCH v2] " Vince Kim
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).