OMAP: iovmm: fix SW flags passed by user
Subsystems (2)
PATCH 1/1 OMAP: iovmm: fix SW flags passed by user
Commit d038aee24dcd changes iovmm to receive flags specified by user, however the upper 16 bits of the flags are wiped by iovmm itself.
This fixes IOVMF_DA_FIXED flags from being lost, and lets the user map its desired "device addresses".
--- arch/arm/plat-omap/include/plat/iovmm.h | 3 --- arch/arm/plat-omap/iovmm.c | 4 ---- 2 files changed, 0 insertions(+), 7 deletions(-)
arch/arm/plat-omap/include/plat/iovmm.h +0 -3
@@ -29,9 +29,6 @@ struct iovm_struct {* lower 16 bit is used for h/w and upper 16 bit is for s/w. */ #define IOVMF_SW_SHIFT 16 -#define IOVMF_HW_SIZE (1 << IOVMF_SW_SHIFT) -#define IOVMF_HW_MASK (IOVMF_HW_SIZE - 1) -#define IOVMF_SW_MASK (~IOVMF_HW_MASK)UL/* * iovma: h/w flags derived from cam and ram attribute
arch/arm/plat-omap/iovmm.c +0 -4
@@ -648,7 +648,6 @@ u32 iommu_vmap(struct iommu *obj, u32 da, const struct sg_table *sgt,return PTR_ERR (va); } - flags &=IOVMF_HW_MASK ; flags |=IOVMF_DISCONT ; flags |=IOVMF_MMIO ;
@@ -706,7 +705,6 @@ u32 iommu_vmalloc(struct iommu *obj, u32 da, size_t bytes, u32 flags)if (!va)return -ENOMEM ; - flags &=IOVMF_HW_MASK ; flags |=IOVMF_DISCONT ; flags |=IOVMF_ALLOC ;
@@ -795,7 +793,6 @@ u32 iommu_kmap(struct iommu *obj, u32 da, u32 pa, size_t bytes,if (!va)return -ENOMEM ; - flags &=IOVMF_HW_MASK ; flags |=IOVMF_LINEAR ; flags |=IOVMF_MMIO ;
@@ -853,7 +850,6 @@ u32 iommu_kmalloc(struct iommu *obj, u32 da, size_t bytes, u32 flags)return -ENOMEM ; pa =virt_to_phys (va); - flags &=IOVMF_HW_MASK ; flags |=IOVMF_LINEAR ; flags |=IOVMF_ALLOC ;
Signature · 1 line
1.7.1
PATCH 1/1 OMAP: iovmm: fix SW flags passed by user
Hello.
Quoted 1 line · Omar Ramirez Luna
Commit
d038aee24dcdchanges iovmm to receive flags specified by user,
Please also specify that commit's summary for the human readers. Bare ID is
only usable to gitweb...
Quoted 6 lines
however the upper 16 bits of the flags are wiped by iovmm itself.
This fixes
IOVMF_DA_FIXEDflags from being lost, and lets the user map its desired "device addresses".Signed-off-by Omar Ramirez Luna <[email protected]>
WBR, Sergei
PATCH 1/1 OMAP: iovmm: fix SW flags passed by user
Quoted 8 lines · Sergei Shtylyov
Hello.
Quoted 1 line · Omar Ramirez Luna
Commit
d038aee24dcdchanges iovmm to receive flags specified by user,? Please also specify that commit's summary for the human readers. Bare ID is only usable to gitweb...
Here it is the patch, introducing the change:
https://patchwork.kernel.org/patch/620871/
If accepted by Hiroshi, I can update the description making reference to this patch instead of commit.
Regards,
Omar
PATCH 1/1 OMAP: iovmm: fix SW flags passed by user
Quoted 10 lines · Omar
Quoted 8 lines · Sergei Shtylyov
Hello.
Quoted 1 line · Omar Ramirez Luna
Commit
d038aee24dcdchanges iovmm to receive flags specified by user,? Please also specify that commit's summary for the human readers. Bare ID is only usable to gitweb...
Quoted 6 lines
Here it is the patch, introducing the change:
https://patchwork.kernel.org/patch/620871/
If accepted by Hiroshi, I can update the description making reference to this patch instead of commit.
It is a requirement that all descriptions which refer to a commit shall contain both the commit ID *and* the summary line for that commit. Failure to ensure that's the case results in rejection of the patch until it conforms.
PATCH 1/1 OMAP: iovmm: fix SW flags passed by user
On Fri, Mar 25, 2011 at 3:05 PM, Russell King - ARM Linux <[email protected]>
Quoted 9 lines
Quoted 9 lines
Quoted 4 lines
Quoted 1 line
Commit
d038aee24dcdchanges iovmm to receive flags specified by user,? Please also specify that commit's summary for the human readers. Bare ID is only usable to gitweb...
Here it is the patch, introducing the change:
Quoted 7 lines
Quoted 2 lines
If accepted by Hiroshi, I can update the description making reference to this patch instead of commit.
It is a requirement that all descriptions which refer to a commit shall contain both the commit ID *and* the summary line for that commit. Failure to ensure that's the case results in rejection of the patch until it conforms.
Not a problem then... resending v2.
Regards,
Omar