Message ID | 20180227231952.GA9001@embeddedor.com (mailing list archive) |
---|---|
State | New, archived |
Delegated to: | Bjorn Helgaas |
Headers | show |
On Wed, Feb 28, 2018 at 1:19 AM, Gustavo A. R. Silva <garsilva@embeddedor.com> wrote: > It seems that the expression threshold_us * 1000 will never exceed the > 32-bit limits [1]. So changing the type of threshold_ns from u64 to u32 > seems sensible [2]. > > [1] https://marc.info/?l=linux-kernel&m=151855021100725&w=2 > [2] https://marc.info/?l=linux-kernel&m=151976318924615&w=2 > Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com> > Addresses-Coverity-ID: 1462501 > Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> > --- > drivers/pci/pcie/aspm.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c > index 57feef2..8633fc4 100644 > --- a/drivers/pci/pcie/aspm.c > +++ b/drivers/pci/pcie/aspm.c > @@ -322,7 +322,7 @@ static u32 calc_l1ss_pwron(struct pci_dev *pdev, u32 scale, u32 val) > > static void encode_l12_threshold(u32 threshold_us, u32 *scale, u32 *value) > { > - u64 threshold_ns = threshold_us * 1000; > + u32 threshold_ns = threshold_us * 1000; > > /* See PCIe r3.1, sec 7.33.3 and sec 6.18 */ > if (threshold_ns < 32) { > -- > 2.7.4 >
On Tue, Feb 27, 2018 at 05:19:52PM -0600, Gustavo A. R. Silva wrote: > It seems that the expression threshold_us * 1000 will never exceed the > 32-bit limits [1]. So changing the type of threshold_ns from u64 to u32 > seems sensible [2]. > > [1] https://marc.info/?l=linux-kernel&m=151855021100725&w=2 > [2] https://marc.info/?l=linux-kernel&m=151976318924615&w=2 > > Addresses-Coverity-ID: 1462501 > Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> Applied with Andy's Reviewed-by to pci/aspm for v4.17, thanks! I included a more detailed analysis in the changelog: PCI/ASPM: Declare threshold_ns as u32, not u64 aspm_calc_l1ss_info() computes l1_2_threshold in microseconds as: l1_2_threshold = 2 + 4 + t_common_mode + t_power_on; where t_common_mode is at most 255us: PCI_L1SS_CAP_CM_RESTORE_TIME 0x0000ff00 <-- 8 bits; <256us and t_power_on is at most 31 * 100us = 3100us: PCI_L1SS_CAP_P_PWR_ON_VALUE 0x00f80000 <-- 5 bits; <32 PCI_L1SS_CAP_P_PWR_ON_SCALE 0x00030000 <-- *2us, *10us, or *100us So l1_2_threshold is at most 2 + 4 + 255 + 3100 = 3361, which means threshold_ns is at most 3361 * 1000 = 3361000, which easily fits in a u32. Declare threshold_ns as u32, not u64. This fixes a Coverity warning. Addresses-Coverity-ID: 1462501 Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> [bhelgaas: changelog] Signed-off-by: Bjorn Helgaas <helgaas@kernel.org> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com> > --- > drivers/pci/pcie/aspm.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c > index 57feef2..8633fc4 100644 > --- a/drivers/pci/pcie/aspm.c > +++ b/drivers/pci/pcie/aspm.c > @@ -322,7 +322,7 @@ static u32 calc_l1ss_pwron(struct pci_dev *pdev, u32 scale, u32 val) > > static void encode_l12_threshold(u32 threshold_us, u32 *scale, u32 *value) > { > - u64 threshold_ns = threshold_us * 1000; > + u32 threshold_ns = threshold_us * 1000; > > /* See PCIe r3.1, sec 7.33.3 and sec 6.18 */ > if (threshold_ns < 32) { > -- > 2.7.4 >
On 02/28/2018 02:55 PM, Bjorn Helgaas wrote: > On Tue, Feb 27, 2018 at 05:19:52PM -0600, Gustavo A. R. Silva wrote: >> It seems that the expression threshold_us * 1000 will never exceed the >> 32-bit limits [1]. So changing the type of threshold_ns from u64 to u32 >> seems sensible [2]. >> >> [1] https://marc.info/?l=linux-kernel&m=151855021100725&w=2 >> [2] https://marc.info/?l=linux-kernel&m=151976318924615&w=2 >> >> Addresses-Coverity-ID: 1462501 >> Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> > > Applied with Andy's Reviewed-by to pci/aspm for v4.17, thanks! > > I included a more detailed analysis in the changelog: > > PCI/ASPM: Declare threshold_ns as u32, not u64 > > aspm_calc_l1ss_info() computes l1_2_threshold in microseconds as: > > l1_2_threshold = 2 + 4 + t_common_mode + t_power_on; > > where t_common_mode is at most 255us: > > PCI_L1SS_CAP_CM_RESTORE_TIME 0x0000ff00 <-- 8 bits; <256us > > and t_power_on is at most 31 * 100us = 3100us: > > PCI_L1SS_CAP_P_PWR_ON_VALUE 0x00f80000 <-- 5 bits; <32 > PCI_L1SS_CAP_P_PWR_ON_SCALE 0x00030000 <-- *2us, *10us, or *100us > > So l1_2_threshold is at most 2 + 4 + 255 + 3100 = 3361, which means > threshold_ns is at most 3361 * 1000 = 3361000, which easily fits in a > u32. > > Declare threshold_ns as u32, not u64. This fixes a Coverity warning. > > Addresses-Coverity-ID: 1462501 > Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> > [bhelgaas: changelog] > Signed-off-by: Bjorn Helgaas <helgaas@kernel.org> Thanks, Bjorn. -- Gustavo > Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com> > >> --- >> drivers/pci/pcie/aspm.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c >> index 57feef2..8633fc4 100644 >> --- a/drivers/pci/pcie/aspm.c >> +++ b/drivers/pci/pcie/aspm.c >> @@ -322,7 +322,7 @@ static u32 calc_l1ss_pwron(struct pci_dev *pdev, u32 scale, u32 val) >> >> static void encode_l12_threshold(u32 threshold_us, u32 *scale, u32 *value) >> { >> - u64 threshold_ns = threshold_us * 1000; >> + u32 threshold_ns = threshold_us * 1000; >> >> /* See PCIe r3.1, sec 7.33.3 and sec 6.18 */ >> if (threshold_ns < 32) { >> -- >> 2.7.4 >>
diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c index 57feef2..8633fc4 100644 --- a/drivers/pci/pcie/aspm.c +++ b/drivers/pci/pcie/aspm.c @@ -322,7 +322,7 @@ static u32 calc_l1ss_pwron(struct pci_dev *pdev, u32 scale, u32 val) static void encode_l12_threshold(u32 threshold_us, u32 *scale, u32 *value) { - u64 threshold_ns = threshold_us * 1000; + u32 threshold_ns = threshold_us * 1000; /* See PCIe r3.1, sec 7.33.3 and sec 6.18 */ if (threshold_ns < 32) {
It seems that the expression threshold_us * 1000 will never exceed the 32-bit limits [1]. So changing the type of threshold_ns from u64 to u32 seems sensible [2]. [1] https://marc.info/?l=linux-kernel&m=151855021100725&w=2 [2] https://marc.info/?l=linux-kernel&m=151976318924615&w=2 Addresses-Coverity-ID: 1462501 Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> --- drivers/pci/pcie/aspm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)