linux/io.h: Add cleanup defination for iounmap() and memunmap()
Frank Li
maintainerPATCH 1/1 linux/io.h: Add cleanup defination for iounmap() and memunmap()
Add DEFINE_FREE for iounmap() and memunmap() to support scope based cleanup.
--- include/linux/io.h | 3 +++ 1 file changed, 3 insertions(+)
include/linux/io.h +3 -0
@@ -163,6 +163,9 @@ enum {void *memremap (resource_size_t offset,size_t size,unsigned long flags);void memunmap (void *addr); +DEFINE_FREE(iounmap, void __iomem *, if (!IS_ERR_OR_NULL(_T)) iounmap(_T)) +DEFINE_FREE(memunmap, void *, if (!IS_ERR_OR_NULL(_T)) memunmap(_T)) +/* * On x86 PAT systems we have memory tracking that keeps track of * the allowed mappings on memory ranges. This tracking works for
Signature · 1 line
2.34.1
Quoted 13 lines · Frank Li
Add
DEFINE_FREEforiounmap()andmemunmap()to support scope based cleanup.Signed-off-by Frank Li <[email protected]>--- include/linux/io.h | 3 +++ 1 file changed, 3 insertions(+)include/linux/io.h +0
-0
include/linux/io.h enum { · line 167
Quoted 5 lines
include/linux/io.h +2
-0@@ -163,6 +163,9 @@ enum {void *memremap (resource_size_t offset,size_t size,unsigned long flags);void memunmap (void *addr); +DEFINE_FREE(iounmap, void __iomem *, if (!IS_ERR_OR_NULL(_T)) iounmap(_T)) +DEFINE_FREE(memunmap, void *, if (!IS_ERR_OR_NULL(_T)) memunmap(_T))
I don't like the use of IS_ERR_OR_NULL(), which tends to indicate a problem in the interface design.
In which cases do you expect to need scope based cleanup on an error pointer here? The only interfaces I see that returns an __iomem error pointer are the devm_* ones, but those have their own cleanup method.
Arnd
Frank Li
maintainerQuoted 17 lines · Arnd Bergmann
Quoted 16 lines · Frank Li
Add
DEFINE_FREEforiounmap()andmemunmap()to support scope based cleanup.Signed-off-by Frank Li <[email protected]>--- include/linux/io.h | 3 +++ 1 file changed, 3 insertions(+)include/linux/io.h +0
-0@@ -163,6 +163,9 @@ enum {void *memremap (resource_size_t offset,size_t size,unsigned long flags);void memunmap (void *addr);
Quoted 5 lines
include/linux/io.h enum { · line 167
Quoted 2 lines
include/linux/io.h +2
-0+DEFINE_FREE(iounmap, void __iomem *, if (!IS_ERR_OR_NULL(_T)) iounmap(_T)) +DEFINE_FREE(memunmap, void *, if (!IS_ERR_OR_NULL(_T)) memunmap(_T))I don't like the use of
IS_ERR_OR_NULL(), which tends to indicate a problem in the interface design.
Just !(_T) ? I just refer kfree()'s implementation.
Quoted 5 lines
In which cases do you expect to need scope based cleanup on an error pointer here? The only interfaces I see that returns an
__iomemerror pointer are the devm_* ones, but those have their own cleanup method.
devm_* can help much especial in probe() function. but scope base cleanup also useful.
Give a existed example:
drivers/clocksource/timer-fsl-ftm.c
static int __init ftm_timer_init(struct device_node *np) {
...
priv->clksrc_base = of_iomap(np, 1);
if (!priv->clksrc_base) {
pr_err("ftm: unable to map source timer registers\n");
goto err_clksrc;
}
...err_clksrc:
iounmap(priv->clkevt_base); }
If use scoped base cleanup it will be simple.ftm_timer_init()
{
base __free(ioumap) = of_iomap(np, 1);
...
priv->clksrc_base = no_free_ptr(base);
return 0;
}drivers/clocksource/arm_arch_timer.c
Some example need map and unmap in function such as:arch_timer_mem_frame_get_cntfrq()
{
base = ioremap(frame->cntbase, frame->size);
...
iounmap(base);
}of course this example is too simple. If there are many error branch between ioremap() and iounmap(), scoped based clean up make code simple.
Frank
Quoted 2 lines
Arnd
Quoted 10 lines · Frank Li
Quoted 9 lines · Arnd Bergmann
Quoted 8 lines · Frank Li
include/linux/io.h +0
-0@@ -163,6 +163,9 @@ enum {void *memremap (resource_size_t offset,size_t size,unsigned long flags);void memunmap (void *addr);
Quoted 7 lines
Quoted 5 lines
include/linux/io.h enum { · line 167
Quoted 2 lines
include/linux/io.h +2
-0+DEFINE_FREE(iounmap, void __iomem *, if (!IS_ERR_OR_NULL(_T)) iounmap(_T)) +DEFINE_FREE(memunmap, void *, if (!IS_ERR_OR_NULL(_T)) memunmap(_T))I don't like the use of
IS_ERR_OR_NULL(), which tends to indicate a problem in the interface design.Just !(_T) ? I just refer kfree()'s implementation.
Right, I think !(_T) is what we want here. I did see that kfree has the same construct you are adding here, but did not read up on how it got there.
Quoted 8 lines
Quoted 4 lines
In which cases do you expect to need scope based cleanup on an error pointer here? The only interfaces I see that returns an
__iomemerror pointer are the devm_* ones, but those have their own cleanup method.devm_* can help much especial in
probe()function. but scope base cleanup also useful.
Yes, that's fine. My only point here is that you can't combine the cleanup function you added with devm_ioremap(), as that needs to be paired with either implicit cleanup or with devm_iounmap() but not plain iounmap().
Arnd