drivers/iommu/riscv/iommu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
From: Guo Ren <guoren@linux.alibaba.com>
When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
than unsigned long. Therefore, we should use size_t to avoid
type-checking errors.
Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
Signed-off-by: Guo Ren <guoren@kernel.org>
Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
---
drivers/iommu/riscv/iommu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/iommu/riscv/iommu.c b/drivers/iommu/riscv/iommu.c
index 8a05def774bd..38d381164385 100644
--- a/drivers/iommu/riscv/iommu.c
+++ b/drivers/iommu/riscv/iommu.c
@@ -1270,7 +1270,7 @@ static phys_addr_t riscv_iommu_iova_to_phys(struct iommu_domain *iommu_domain,
dma_addr_t iova)
{
struct riscv_iommu_domain *domain = iommu_domain_to_riscv(iommu_domain);
- unsigned long pte_size;
+ size_t pte_size;
unsigned long *ptr;
ptr = riscv_iommu_pte_fetch(domain, iova, &pte_size);
--
2.40.1
On Thu, Jan 02, 2025 at 09:46:16PM -0500, guoren@kernel.org wrote:
> From: Guo Ren <guoren@linux.alibaba.com>
>
> When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
> than unsigned long. Therefore, we should use size_t to avoid
> type-checking errors.
>
> Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
> Signed-off-by: Guo Ren <guoren@kernel.org>
> Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
> ---
> drivers/iommu/riscv/iommu.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
Applied, thanks.
On Thu, Jan 02, 2025 at 09:46:16PM -0500, guoren@kernel.org wrote:
> From: Guo Ren <guoren@linux.alibaba.com>
>
> When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
> than unsigned long. Therefore, we should use size_t to avoid
> type-checking errors.
>
> Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
> Signed-off-by: Guo Ren <guoren@kernel.org>
> Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
> ---
> drivers/iommu/riscv/iommu.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iommu/riscv/iommu.c b/drivers/iommu/riscv/iommu.c
> index 8a05def774bd..38d381164385 100644
> --- a/drivers/iommu/riscv/iommu.c
> +++ b/drivers/iommu/riscv/iommu.c
> @@ -1270,7 +1270,7 @@ static phys_addr_t riscv_iommu_iova_to_phys(struct iommu_domain *iommu_domain,
> dma_addr_t iova)
> {
> struct riscv_iommu_domain *domain = iommu_domain_to_riscv(iommu_domain);
> - unsigned long pte_size;
> + size_t pte_size;
riscv_iommu_pte_fetch() expects size_t so even though ILP32 isn't
currently supported in the kernel, this change makes sense.
Reviewed-by: Charlie Jenkins <charlie@rivosinc.com>
> unsigned long *ptr;
>
> ptr = riscv_iommu_pte_fetch(domain, iova, &pte_size);
> --
> 2.40.1
>
>
> _______________________________________________
> linux-riscv mailing list
> linux-riscv@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv
Hi Guo,
On 03/01/2025 03:46, guoren@kernel.org wrote:
> From: Guo Ren <guoren@linux.alibaba.com>
>
> When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
RISCV_IOMMU depends on 64BIT so how do you get __BITS_PER_LONG == 32?
Thanks,
Alex
> than unsigned long. Therefore, we should use size_t to avoid
> type-checking errors.
>
> Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
> Signed-off-by: Guo Ren <guoren@kernel.org>
> Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
> ---
> drivers/iommu/riscv/iommu.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/iommu/riscv/iommu.c b/drivers/iommu/riscv/iommu.c
> index 8a05def774bd..38d381164385 100644
> --- a/drivers/iommu/riscv/iommu.c
> +++ b/drivers/iommu/riscv/iommu.c
> @@ -1270,7 +1270,7 @@ static phys_addr_t riscv_iommu_iova_to_phys(struct iommu_domain *iommu_domain,
> dma_addr_t iova)
> {
> struct riscv_iommu_domain *domain = iommu_domain_to_riscv(iommu_domain);
> - unsigned long pte_size;
> + size_t pte_size;
> unsigned long *ptr;
>
> ptr = riscv_iommu_pte_fetch(domain, iova, &pte_size);
Hi Alex,
On 2025-01-06 4:31 AM, Alexandre Ghiti wrote:
> On 03/01/2025 03:46, guoren@kernel.org wrote:
>> From: Guo Ren <guoren@linux.alibaba.com>
>>
>> When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
>
>
> RISCV_IOMMU depends on 64BIT so how do you get __BITS_PER_LONG == 32?
I am guessing this is from some downstream effort to compile a 64-bit kernel
with an ILP32 ABI. However unsupported that may be, size_t is the nominal type
used in both riscv_iommu_pte_fetch() and above in iommu_iotlb_gather_add_page(),
so the change does make sense.
Regards,
Samuel
>> than unsigned long. Therefore, we should use size_t to avoid
>> type-checking errors.
>>
>> Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
>> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
>> Signed-off-by: Guo Ren <guoren@kernel.org>
>> Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
>> ---
>> drivers/iommu/riscv/iommu.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/iommu/riscv/iommu.c b/drivers/iommu/riscv/iommu.c
>> index 8a05def774bd..38d381164385 100644
>> --- a/drivers/iommu/riscv/iommu.c
>> +++ b/drivers/iommu/riscv/iommu.c
>> @@ -1270,7 +1270,7 @@ static phys_addr_t riscv_iommu_iova_to_phys(struct
>> iommu_domain *iommu_domain,
>> dma_addr_t iova)
>> {
>> struct riscv_iommu_domain *domain = iommu_domain_to_riscv(iommu_domain);
>> - unsigned long pte_size;
>> + size_t pte_size;
>> unsigned long *ptr;
>> ptr = riscv_iommu_pte_fetch(domain, iova, &pte_size);
Hi Samuel & Alexandre,
On Tue, Jan 7, 2025 at 1:47 AM Samuel Holland <samuel.holland@sifive.com> wrote:
>
> Hi Alex,
>
> On 2025-01-06 4:31 AM, Alexandre Ghiti wrote:
> > On 03/01/2025 03:46, guoren@kernel.org wrote:
> >> From: Guo Ren <guoren@linux.alibaba.com>
> >>
> >> When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
> >
> >
> > RISCV_IOMMU depends on 64BIT so how do you get __BITS_PER_LONG == 32?
>
> I am guessing this is from some downstream effort to compile a 64-bit kernel
> with an ILP32 ABI. However unsupported that may be, size_t is the nominal type
> used in both riscv_iommu_pte_fetch() and above in iommu_iotlb_gather_add_page(),
> so the change does make sense.
Yes, I'm working on ILP32 ABI on CONFIG_64BIT=y. So, I got the warning
when __BITS_PER_LONG == 32.
In riscv_iommu_unmap_pages(), we've written "size_t pte_size;." So, we
should do that in "riscv_iommu_iova_to_phys()".
>
> Regards,
> Samuel
>
> >> than unsigned long. Therefore, we should use size_t to avoid
> >> type-checking errors.
> >>
> >> Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
> >> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
> >> Signed-off-by: Guo Ren <guoren@kernel.org>
> >> Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
> >> ---
> >> drivers/iommu/riscv/iommu.c | 2 +-
> >> 1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/iommu/riscv/iommu.c b/drivers/iommu/riscv/iommu.c
> >> index 8a05def774bd..38d381164385 100644
> >> --- a/drivers/iommu/riscv/iommu.c
> >> +++ b/drivers/iommu/riscv/iommu.c
> >> @@ -1270,7 +1270,7 @@ static phys_addr_t riscv_iommu_iova_to_phys(struct
> >> iommu_domain *iommu_domain,
> >> dma_addr_t iova)
> >> {
> >> struct riscv_iommu_domain *domain = iommu_domain_to_riscv(iommu_domain);
> >> - unsigned long pte_size;
> >> + size_t pte_size;
> >> unsigned long *ptr;
> >> ptr = riscv_iommu_pte_fetch(domain, iova, &pte_size);
>
--
Best Regards
Guo Ren
On Tue, Jan 7, 2025 at 10:48 PM Guo Ren <guoren@kernel.org> wrote:
>
> Hi Samuel & Alexandre,
>
> On Tue, Jan 7, 2025 at 1:47 AM Samuel Holland <samuel.holland@sifive.com> wrote:
> >
> > Hi Alex,
> >
> > On 2025-01-06 4:31 AM, Alexandre Ghiti wrote:
> > > On 03/01/2025 03:46, guoren@kernel.org wrote:
> > >> From: Guo Ren <guoren@linux.alibaba.com>
> > >>
> > >> When __BITS_PER_LONG == 32, size_t is defined as unsigned int rather
> > >
> > >
> > > RISCV_IOMMU depends on 64BIT so how do you get __BITS_PER_LONG == 32?
> >
> > I am guessing this is from some downstream effort to compile a 64-bit kernel
> > with an ILP32 ABI. However unsupported that may be, size_t is the nominal type
> > used in both riscv_iommu_pte_fetch() and above in iommu_iotlb_gather_add_page(),
> > so the change does make sense.
> Yes, I'm working on ILP32 ABI on CONFIG_64BIT=y. So, I got the warning
> when __BITS_PER_LONG == 32.
>
> In riscv_iommu_unmap_pages(), we've written "size_t pte_size;." So, we
> should do that in "riscv_iommu_iova_to_phys()".
>
Thanks for pointing this out.
Do we have to change other relevant uses of `unsigned long` to either
`phys_addr_t` or `dma_addr_t` to avoid type mismatch and get ILP32
working correctly?
Anyway, for quick warning fix:
Reviewed-by: Tomasz Jeznach <tjeznach@rivosinc.com>
> >
> > Regards,
> > Samuel
> >
> > >> than unsigned long. Therefore, we should use size_t to avoid
> > >> type-checking errors.
> > >>
> > >> Fixes: 488ffbf18171 ("iommu/riscv: Paging domain support")
> > >> Signed-off-by: Guo Ren <guoren@linux.alibaba.com>
> > >> Signed-off-by: Guo Ren <guoren@kernel.org>
> > >> Cc: Tomasz Jeznach <tjeznach@rivosinc.com>
> > >> ---
> > >> drivers/iommu/riscv/iommu.c | 2 +-
> > >> 1 file changed, 1 insertion(+), 1 deletion(-)
> > >>
> > >> diff --git a/drivers/iommu/riscv/iommu.c b/drivers/iommu/riscv/iommu.c
> > >> index 8a05def774bd..38d381164385 100644
> > >> --- a/drivers/iommu/riscv/iommu.c
> > >> +++ b/drivers/iommu/riscv/iommu.c
> > >> @@ -1270,7 +1270,7 @@ static phys_addr_t riscv_iommu_iova_to_phys(struct
> > >> iommu_domain *iommu_domain,
> > >> dma_addr_t iova)
> > >> {
> > >> struct riscv_iommu_domain *domain = iommu_domain_to_riscv(iommu_domain);
> > >> - unsigned long pte_size;
> > >> + size_t pte_size;
> > >> unsigned long *ptr;
> > >> ptr = riscv_iommu_pte_fetch(domain, iova, &pte_size);
> >
>
>
> --
> Best Regards
> Guo Ren
© 2016 - 2026 Red Hat, Inc.