Re: [PATCH AUTOSEL 5.15 11/19] usb: gadget: fsl_qe_udc: validate endpoint index for ch9 udc
On 9/9/23 3:38 AM, Sasha Levin wrote:
> From: Ma Ke
>
> [ Upstream commit ce9daa2efc0872a9a68ea51dc8000df05893ef2e ]
>
> We should verify the bound of the array to assure that host
> may not manipulate the index to point past endpoint array.
>
> Signed-off-by: Ma Ke
> Acked-by: Li Yang
> Link: https://lore.kernel.org/r/20230628081511.186850-1-make_ruc2...@163.com
> Signed-off-by: Greg Kroah-Hartman
> Signed-off-by: Sasha Levin
> ---
> drivers/usb/gadget/udc/fsl_qe_udc.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/usb/gadget/udc/fsl_qe_udc.c
> b/drivers/usb/gadget/udc/fsl_qe_udc.c
> index 15db7a3868fe4..aff4050f96dd6 100644
> --- a/drivers/usb/gadget/udc/fsl_qe_udc.c
> +++ b/drivers/usb/gadget/udc/fsl_qe_udc.c
> @@ -1956,6 +1956,8 @@ static void ch9getstatus(struct qe_udc *udc, u8
> request_type, u16 value,
> } else if ((request_type & USB_RECIP_MASK) == USB_RECIP_ENDPOINT) {
> /* Get endpoint status */
> int pipe = index & USB_ENDPOINT_NUMBER_MASK;
> + if (pipe >= USB_MAX_ENDPOINTS)
> + goto stall;
Hm, what's the earliest version of gcc needed to compile this
(*if* statement amongst the variable declarations)?
> struct qe_ep *target_ep = &udc->eps[pipe];
> u16 usep;
>
MBR, Sergey
Re: [PATCH 00/34] biops: add atomig find_bit() operations
On 11/18/23 7:18 PM, Bart Van Assche wrote:
[...]
>> Add helpers around test_and_{set,clear}_bit() that allow to search for
>> clear or set bits and flip them atomically.
>
> There is a typo in the subject: shouldn't "atomig" be changed
> into "atomic"?
And "biops" to "bitops"? :-)
> Thanks,
>
> Bart.
MBR, Sergey
Re: [PATCH 7/9] ata: pata_falcon: Remove #include
On 8/17/23 7:07 PM, Geert Uytterhoeven wrote: > The Atari Falcon PATA driver does not need anything from . > > Signed-off-by: Geert Uytterhoeven Reviewed-by: Sergey Shtylyov [...] MBR, Sergey
Re: [PATCH 8/9] ata: pata_gayle: Remove #include
On 8/17/23 7:07 PM, Geert Uytterhoeven wrote: > The Amiga Gayle PATA driver does not need anything from . > > Signed-off-by: Geert Uytterhoeven Reviewed-by: Sergey Shtylyov [...] MBR, Sergey
Re: [PATCH 6/9] ata: pata_buddha: Remove #include
On 8/17/23 7:07 PM, Geert Uytterhoeven wrote: > The Buddha, Catweasel, and X-Surf PATA driver does not need anything > from . > > Signed-off-by: Geert Uytterhoeven Reviewed-by: Sergey Shtylyov [...] MBR, Sergey
[PATCH 00/13] Make user_regset_copyin_ignore() *void*
Here are 13 patches against the 'next-20221014' tag of the 'linux-next.git' repo. I'm not sure how this cross-arch series should be merged -- perhaps thru Andrew Morton's tree? user_regset_copyin_ignore() apparently cannot fail and so always returns 0. Let's first remove the result checks in several architectures that call this function and then make user_regset_copyin_ignore() return *void* instead of *int*... Sergey Shtylyov (13): arc: ptrace: user_regset_copyin_ignore() always returns 0 arm: ptrace: user_regset_copyin_ignore() always returns 0 arm64: ptrace: user_regset_copyin_ignore() always returns 0 hexagon: ptrace: user_regset_copyin_ignore() always returns 0 ia64: ptrace: user_regset_copyin_ignore() always returns 0 mips: ptrace: user_regset_copyin_ignore() always returns 0 nios2: ptrace: user_regset_copyin_ignore() always returns 0 openrisc: ptrace: user_regset_copyin_ignore() always returns 0 parisc: ptrace: user_regset_copyin_ignore() always returns 0 powerpc: ptrace: user_regset_copyin_ignore() always returns 0 sh: ptrace: user_regset_copyin_ignore() always returns 0 sparc: ptrace: user_regset_copyin_ignore() always returns 0 regset: make user_regset_copyin_ignore() *void* arch/arc/kernel/ptrace.c | 2 +- arch/arm/kernel/ptrace.c | 8 +++- arch/arm64/kernel/ptrace.c | 16 arch/hexagon/kernel/ptrace.c | 7 +++ arch/ia64/kernel/ptrace.c| 20 +--- arch/mips/kernel/ptrace.c| 9 + arch/nios2/kernel/ptrace.c | 6 +++--- arch/openrisc/kernel/ptrace.c| 8 +++- arch/parisc/kernel/ptrace.c | 15 +-- arch/powerpc/kernel/ptrace/ptrace-tm.c | 10 +- arch/powerpc/kernel/ptrace/ptrace-view.c | 10 +- arch/sh/kernel/ptrace_32.c | 8 arch/sparc/kernel/ptrace_32.c| 9 + arch/sparc/kernel/ptrace_64.c| 23 +++ include/linux/regset.h | 15 +++ 15 files changed, 77 insertions(+), 89 deletions(-) -- 2.26.3
[PATCH 10/13] powerpc: ptrace: user_regset_copyin_ignore() always returns 0
user_regset_copyin_ignore() always returns 0, so checking its result seems
pointless -- don't do this anymore...
Signed-off-by: Sergey Shtylyov
---
arch/powerpc/kernel/ptrace/ptrace-tm.c | 10 +-
arch/powerpc/kernel/ptrace/ptrace-view.c | 10 +-
2 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/arch/powerpc/kernel/ptrace/ptrace-tm.c
b/arch/powerpc/kernel/ptrace/ptrace-tm.c
index 44045363a903..210ea834e603 100644
--- a/arch/powerpc/kernel/ptrace/ptrace-tm.c
+++ b/arch/powerpc/kernel/ptrace/ptrace-tm.c
@@ -170,9 +170,9 @@ int tm_cgpr_set(struct task_struct *target, const struct
user_regset *regset,
(PT_MAX_PUT_REG + 1) * sizeof(reg));
if (PT_MAX_PUT_REG + 1 < PT_TRAP && !ret)
- ret = user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
- (PT_MAX_PUT_REG + 1) *
sizeof(reg),
- PT_TRAP * sizeof(reg));
+ user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
+ (PT_MAX_PUT_REG + 1) * sizeof(reg),
+ PT_TRAP * sizeof(reg));
if (!ret && count > 0) {
ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, ®,
@@ -183,8 +183,8 @@ int tm_cgpr_set(struct task_struct *target, const struct
user_regset *regset,
}
if (!ret)
- ret = user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
- (PT_TRAP + 1) * sizeof(reg),
-1);
+ user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
+ (PT_TRAP + 1) * sizeof(reg), -1);
return ret;
}
diff --git a/arch/powerpc/kernel/ptrace/ptrace-view.c
b/arch/powerpc/kernel/ptrace/ptrace-view.c
index 076d867412c7..ca0bf8da48fd 100644
--- a/arch/powerpc/kernel/ptrace/ptrace-view.c
+++ b/arch/powerpc/kernel/ptrace/ptrace-view.c
@@ -267,9 +267,9 @@ static int gpr_set(struct task_struct *target, const struct
user_regset *regset,
(PT_MAX_PUT_REG + 1) * sizeof(reg));
if (PT_MAX_PUT_REG + 1 < PT_TRAP && !ret)
- ret = user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
- (PT_MAX_PUT_REG + 1) *
sizeof(reg),
- PT_TRAP * sizeof(reg));
+ user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
+ (PT_MAX_PUT_REG + 1) * sizeof(reg),
+ PT_TRAP * sizeof(reg));
if (!ret && count > 0) {
ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, ®,
@@ -280,8 +280,8 @@ static int gpr_set(struct task_struct *target, const struct
user_regset *regset,
}
if (!ret)
- ret = user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
- (PT_TRAP + 1) * sizeof(reg),
-1);
+ user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
+ (PT_TRAP + 1) * sizeof(reg), -1);
return ret;
}
--
2.26.3
Re: [PATCH v3 10/24] sparc: Remove COMMAND_LINE_SIZE from uapi
On 2/14/23 10:49 AM, Alexandre Ghiti wrote: > From: Palmer Dabbelt > > As far as I can tell this is not used by userspace and thus should not > be part of the user-visible API. > > Signed-off-by: Palmer Dabbelt > --- > arch/sparc/include/asm/setup.h | 6 +- > arch/sparc/include/uapi/asm/setup.h | 7 --- > 2 files changed, 5 insertions(+), 8 deletions(-) > > diff --git a/arch/sparc/include/asm/setup.h b/arch/sparc/include/asm/setup.h > index 72205684e51e..d1384ed92547 100644 > --- a/arch/sparc/include/asm/setup.h > +++ b/arch/sparc/include/asm/setup.h > @@ -7,7 +7,11 @@ > > #include > > -#include > +#if defined(__sparc__) && defined(__arch64__) Mhm, I don't think these two can be #define'd simulaneously... > +# define COMMAND_LINE_SIZE 2048 > +#else > +# define COMMAND_LINE_SIZE 256 > +#endif > > extern char reboot_command[]; > > diff --git a/arch/sparc/include/uapi/asm/setup.h > b/arch/sparc/include/uapi/asm/setup.h > index 3c208a4dd464..c3cf1b0d30b3 100644 > --- a/arch/sparc/include/uapi/asm/setup.h > +++ b/arch/sparc/include/uapi/asm/setup.h > @@ -6,11 +6,4 @@ > #ifndef _UAPI_SPARC_SETUP_H > #define _UAPI_SPARC_SETUP_H > > -#if defined(__sparc__) && defined(__arch64__) > -# define COMMAND_LINE_SIZE 2048 > -#else > -# define COMMAND_LINE_SIZE 256 > -#endif > - > - > #endif /* _UAPI_SPARC_SETUP_H */ MBR, Sergey
Re: [PATCH v3 10/24] sparc: Remove COMMAND_LINE_SIZE from uapi
On 2/14/23 11:59 AM, WANG Xuerui wrote: [...] >>> From: Palmer Dabbelt >>> >>> As far as I can tell this is not used by userspace and thus should not >>> be part of the user-visible API. >>> >>> Signed-off-by: Palmer Dabbelt >>> --- >>> arch/sparc/include/asm/setup.h | 6 +- >>> arch/sparc/include/uapi/asm/setup.h | 7 --- >>> 2 files changed, 5 insertions(+), 8 deletions(-) >>> >>> diff --git a/arch/sparc/include/asm/setup.h b/arch/sparc/include/asm/setup.h >>> index 72205684e51e..d1384ed92547 100644 >>> --- a/arch/sparc/include/asm/setup.h >>> +++ b/arch/sparc/include/asm/setup.h >>> @@ -7,7 +7,11 @@ >>> #include >>> -#include >>> +#if defined(__sparc__) && defined(__arch64__) >> >> Mhm, I don't think these two can be #define'd simulaneously... > > I believe it's just a SPARC-ism [1] [2] that may look strange and be easily > confused for __aarch64__ (notice the extra 'a')... Yeah, sorry, I did confuse it with AArgh64. :-) [...] MNR, Sergey
[PATCH v2 7/9] usb: phy: fsl-usb: add IRQ check
The driver neglects to check the result of platform_get_irq()'s call and
blithely passes the negative error codes to request_irq() (which takes
*unsigned* IRQ #), causing it to fail with -EINVAL, overriding an original
error code. Stop calling request_irq() with the invalid IRQ #s.
Fixes: 0807c500a1a6 ("USB: add Freescale USB OTG Transceiver driver")
Signed-off-by: Sergey Shtylyov
---
drivers/usb/phy/phy-fsl-usb.c |2 ++
1 file changed, 2 insertions(+)
Index: usb/drivers/usb/phy/phy-fsl-usb.c
===
--- usb.orig/drivers/usb/phy/phy-fsl-usb.c
+++ usb/drivers/usb/phy/phy-fsl-usb.c
@@ -873,6 +873,8 @@ int usb_otg_start(struct platform_device
/* request irq */
p_otg->irq = platform_get_irq(pdev, 0);
+ if (p_otg->irq < 0)
+ return p_otg->irq;
status = request_irq(p_otg->irq, fsl_otg_isr,
IRQF_SHARED, driver_name, p_otg);
if (status) {
Re: consolidate the compat fcntl definitions
Hello! On 1/11/22 11:35 AM, Christoph Hellwig wrote: > currenty the compat fcnt definitions are duplicate for all compat fcntl? > architectures, and the native fcntl64 definitions aren't even usable > from userspace due to a bogus CONFIG_64BIT ifdef. This series tries > to sort out all that. [...] MBR, Sergey
