| Message ID | 54EC38CA.6050607@plexistor.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 02/24/2015 10:39 AM, Boaz Harrosh wrote: > > Resource providers set this flag if they want > that request_region_XXX will print a warning in dmesg > if this particular resource is locked by a driver. > > Thous acting as a Protocol Police about experimental > devices that did not pass a comity approval. > > The Only user of this flag is x86/kernel/e820.c that > wants to WARN about UNKNOWN memory types. > Hi Ingo So I slept on it and I think this simple version is safe and does what we need. It is more simple than the wrappers thing Who can push such a patch, can you push it through your tree or is there another maintainer that needs to push this? Who can we ask about the safeness of these flags? > Signed-off-by: Boaz Harrosh <boaz@plexistor.com> <> Thanks Boaz
* Boaz Harrosh <boaz@plexistor.com> wrote: > --- a/include/linux/ioport.h > +++ b/include/linux/ioport.h > @@ -255,6 +256,5 @@ static inline bool resource_overlaps(struct resource *r1, struct resource *r2) > return (r1->start <= r2->end && r1->end >= r2->start); > } > > - > #endif /* __ASSEMBLY__ */ While the newline is spurious, it should probably not be removed in this patch. Thanks, Ingo
* Boaz Harrosh <boaz@plexistor.com> wrote: > --- a/kernel/resource.c > +++ b/kernel/resource.c > @@ -1075,8 +1075,13 @@ struct resource * __request_region(struct resource *parent, > break; > if (conflict != parent) { > parent = conflict; > - if (!(conflict->flags & IORESOURCE_BUSY)) > + if (!(conflict->flags & IORESOURCE_BUSY)) { > + if (conflict->flags & IORESOURCE_MEM_WARN) > + pr_warn("request unknown region [mem %#010llx-%#010llx] %s\n", > + conflict->start, conflict->end, > + conflict->name); Maybe: 'request region with unknown memory type [mem ...]'? The driver (we hope!) very well knows about the region so it's not totally unknown. Thanks, Ingo
On 02/24/2015 11:06 AM, Ingo Molnar wrote: > > * Boaz Harrosh <boaz@plexistor.com> wrote: > >> --- a/include/linux/ioport.h >> +++ b/include/linux/ioport.h >> @@ -255,6 +256,5 @@ static inline bool resource_overlaps(struct resource *r1, struct resource *r2) >> return (r1->start <= r2->end && r1->end >= r2->start); >> } >> >> - >> #endif /* __ASSEMBLY__ */ > > While the newline is spurious, it should probably not be > removed in this patch. > Rrrr you are right. thanks. I did not notice that I did this. I will resend. Boaz
On 02/24/2015 11:07 AM, Ingo Molnar wrote: > > * Boaz Harrosh <boaz@plexistor.com> wrote: > >> --- a/kernel/resource.c >> +++ b/kernel/resource.c >> @@ -1075,8 +1075,13 @@ struct resource * __request_region(struct resource *parent, >> break; >> if (conflict != parent) { >> parent = conflict; >> - if (!(conflict->flags & IORESOURCE_BUSY)) >> + if (!(conflict->flags & IORESOURCE_BUSY)) { >> + if (conflict->flags & IORESOURCE_MEM_WARN) >> + pr_warn("request unknown region [mem %#010llx-%#010llx] %s\n", >> + conflict->start, conflict->end, >> + conflict->name); > > Maybe: > > 'request region with unknown memory type [mem ...]'? > > The driver (we hope!) very well knows about the region so > it's not totally unknown. > OK makes sense I will fix that too > Thanks, > > Ingo >
diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 1a8a1c3..105bb37 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -961,6 +961,9 @@ void __init e820_reserve_resources(void) res->flags = IORESOURCE_MEM; + if (_is_unknown_type(e820.map[i].type)) + res->flags |= IORESOURCE_MEM_WARN; + /* * don't register the region that could be conflicted with * pci device BAR resource and insert them later in diff --git a/include/linux/ioport.h b/include/linux/ioport.h index 2c525022..183af93 100644 --- a/include/linux/ioport.h +++ b/include/linux/ioport.h @@ -90,6 +90,7 @@ struct resource { #define IORESOURCE_MEM_32BIT (3<<3) #define IORESOURCE_MEM_SHADOWABLE (1<<5) /* dup: IORESOURCE_SHADOWABLE */ #define IORESOURCE_MEM_EXPANSIONROM (1<<6) +#define IORESOURCE_MEM_WARN (1<<7) /* WARN if requested by driver */ /* PnP I/O specific bits (IORESOURCE_BITS) */ #define IORESOURCE_IO_16BIT_ADDR (1<<0) @@ -255,6 +256,5 @@ static inline bool resource_overlaps(struct resource *r1, struct resource *r2) return (r1->start <= r2->end && r1->end >= r2->start); } - #endif /* __ASSEMBLY__ */ #endif /* _LINUX_IOPORT_H */ diff --git a/kernel/resource.c b/kernel/resource.c index 19f2357..fca23ff 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -1075,8 +1075,13 @@ struct resource * __request_region(struct resource *parent, break; if (conflict != parent) { parent = conflict; - if (!(conflict->flags & IORESOURCE_BUSY)) + if (!(conflict->flags & IORESOURCE_BUSY)) { + if (conflict->flags & IORESOURCE_MEM_WARN) + pr_warn("request unknown region [mem %#010llx-%#010llx] %s\n", + conflict->start, conflict->end, + conflict->name); continue; + } } if (conflict->flags & flags & IORESOURCE_MUXED) { add_wait_queue(&muxed_resource_wait, &wait);
Resource providers set this flag if they want that request_region_XXX will print a warning in dmesg if this particular resource is locked by a driver. Thous acting as a Protocol Police about experimental devices that did not pass a comity approval. The Only user of this flag is x86/kernel/e820.c that wants to WARN about UNKNOWN memory types. Signed-off-by: Boaz Harrosh <boaz@plexistor.com> --- arch/x86/kernel/e820.c | 3 +++ include/linux/ioport.h | 2 +- kernel/resource.c | 7 ++++++- 3 files changed, 10 insertions(+), 2 deletions(-)