| Message ID | 20190529122657.166148-3-mimu@linux.ibm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | s390: virtio: support protected virtualization | expand |
On Wed, 29 May 2019 14:26:51 +0200 Michael Mueller <mimu@linux.ibm.com> wrote: > From: Halil Pasic <pasic@linux.ibm.com> > > To support protected virtualization cio will need to make sure the > memory used for communication with the hypervisor is DMA memory. > > Let us introduce one global pool for cio. > > Our DMA pools are implemented as a gen_pool backed with DMA pages. The > idea is to avoid each allocation effectively wasting a page, as we > typically allocate much less than PAGE_SIZE. > > Signed-off-by: Halil Pasic <pasic@linux.ibm.com> > Reviewed-by: Sebastian Ott <sebott@linux.ibm.com> > Signed-off-by: Michael Mueller <mimu@linux.ibm.com> > --- > arch/s390/Kconfig | 1 + > arch/s390/include/asm/cio.h | 11 ++++ > drivers/s390/cio/css.c | 120 ++++++++++++++++++++++++++++++++++++++++++-- > 3 files changed, 128 insertions(+), 4 deletions(-) (...) > diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h > index 1727180e8ca1..43c007d2775a 100644 > --- a/arch/s390/include/asm/cio.h > +++ b/arch/s390/include/asm/cio.h > @@ -328,6 +328,17 @@ static inline u8 pathmask_to_pos(u8 mask) > void channel_subsystem_reinit(void); > extern void css_schedule_reprobe(void); > > +extern void *cio_dma_zalloc(size_t size); > +extern void cio_dma_free(void *cpu_addr, size_t size); > +extern struct device *cio_get_dma_css_dev(void); > + > +struct gen_pool; That forward declaration is a bit ugly... I guess the alternative was include hell? > +void *cio_gp_dma_zalloc(struct gen_pool *gp_dma, struct device *dma_dev, > + size_t size); > +void cio_gp_dma_free(struct gen_pool *gp_dma, void *cpu_addr, size_t size); > +void cio_gp_dma_destroy(struct gen_pool *gp_dma, struct device *dma_dev); > +struct gen_pool *cio_gp_dma_create(struct device *dma_dev, int nr_pages); > + > /* Function from drivers/s390/cio/chsc.c */ > int chsc_sstpc(void *page, unsigned int op, u16 ctrl, u64 *clock_delta); > int chsc_sstpi(void *page, void *result, size_t size); > diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c > index aea502922646..b97618497848 100644 > --- a/drivers/s390/cio/css.c > +++ b/drivers/s390/cio/css.c > @@ -20,6 +20,8 @@ > #include <linux/reboot.h> > #include <linux/suspend.h> > #include <linux/proc_fs.h> > +#include <linux/genalloc.h> > +#include <linux/dma-mapping.h> > #include <asm/isc.h> > #include <asm/crw.h> > > @@ -224,6 +226,8 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid, > INIT_WORK(&sch->todo_work, css_sch_todo); > sch->dev.release = &css_subchannel_release; > device_initialize(&sch->dev); It might be helpful to add a comment why you use 31 bit here... > + sch->dev.coherent_dma_mask = DMA_BIT_MASK(31); > + sch->dev.dma_mask = &sch->dev.coherent_dma_mask; > return sch; > > err: > @@ -899,6 +903,8 @@ static int __init setup_css(int nr) > dev_set_name(&css->device, "css%x", nr); > css->device.groups = cssdev_attr_groups; > css->device.release = channel_subsystem_release; ...and 64 bit here. > + css->device.coherent_dma_mask = DMA_BIT_MASK(64); > + css->device.dma_mask = &css->device.coherent_dma_mask; > > mutex_init(&css->mutex); > css->cssid = chsc_get_cssid(nr); (...) > @@ -1059,16 +1168,19 @@ static int __init css_bus_init(void) > if (ret) > goto out_unregister; > ret = register_pm_notifier(&css_power_notifier); > - if (ret) { > - unregister_reboot_notifier(&css_reboot_notifier); > - goto out_unregister; > - } > + if (ret) > + goto out_unregister_rn; > + ret = cio_dma_pool_init(); > + if (ret) > + goto out_unregister_rn; Don't you also need to unregister the pm notifier on failure here? Other than that, I noticed only cosmetic issues; seems reasonable to me. > css_init_done = 1; > > /* Enable default isc for I/O subchannels. */ > isc_register(IO_SCH_ISC); > > return 0; > +out_unregister_rn: > + unregister_reboot_notifier(&css_reboot_notifier); > out_unregister: > while (i-- > 0) { > struct channel_subsystem *css = channel_subsystems[i];
On 03.06.19 13:37, Cornelia Huck wrote: > On Wed, 29 May 2019 14:26:51 +0200 > Michael Mueller <mimu@linux.ibm.com> wrote: > >> From: Halil Pasic <pasic@linux.ibm.com> >> >> To support protected virtualization cio will need to make sure the >> memory used for communication with the hypervisor is DMA memory. >> >> Let us introduce one global pool for cio. >> >> Our DMA pools are implemented as a gen_pool backed with DMA pages. The >> idea is to avoid each allocation effectively wasting a page, as we >> typically allocate much less than PAGE_SIZE. >> >> Signed-off-by: Halil Pasic <pasic@linux.ibm.com> >> Reviewed-by: Sebastian Ott <sebott@linux.ibm.com> >> Signed-off-by: Michael Mueller <mimu@linux.ibm.com> >> --- >> arch/s390/Kconfig | 1 + >> arch/s390/include/asm/cio.h | 11 ++++ >> drivers/s390/cio/css.c | 120 ++++++++++++++++++++++++++++++++++++++++++-- >> 3 files changed, 128 insertions(+), 4 deletions(-) > > (...) > >> diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h >> index 1727180e8ca1..43c007d2775a 100644 >> --- a/arch/s390/include/asm/cio.h >> +++ b/arch/s390/include/asm/cio.h >> @@ -328,6 +328,17 @@ static inline u8 pathmask_to_pos(u8 mask) >> void channel_subsystem_reinit(void); >> extern void css_schedule_reprobe(void); >> >> +extern void *cio_dma_zalloc(size_t size); >> +extern void cio_dma_free(void *cpu_addr, size_t size); >> +extern struct device *cio_get_dma_css_dev(void); >> + >> +struct gen_pool; > > That forward declaration is a bit ugly... I guess the alternative was > include hell? That's an easy one. #include <linux/bitops.h> +#include <linux/genalloc.h> #include <asm/types.h> > >> +void *cio_gp_dma_zalloc(struct gen_pool *gp_dma, struct device *dma_dev, >> + size_t size); >> +void cio_gp_dma_free(struct gen_pool *gp_dma, void *cpu_addr, size_t size); >> +void cio_gp_dma_destroy(struct gen_pool *gp_dma, struct device *dma_dev); >> +struct gen_pool *cio_gp_dma_create(struct device *dma_dev, int nr_pages); >> + >> /* Function from drivers/s390/cio/chsc.c */ >> int chsc_sstpc(void *page, unsigned int op, u16 ctrl, u64 *clock_delta); >> int chsc_sstpi(void *page, void *result, size_t size); >> diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c >> index aea502922646..b97618497848 100644 >> --- a/drivers/s390/cio/css.c >> +++ b/drivers/s390/cio/css.c >> @@ -20,6 +20,8 @@ >> #include <linux/reboot.h> >> #include <linux/suspend.h> >> #include <linux/proc_fs.h> >> +#include <linux/genalloc.h> >> +#include <linux/dma-mapping.h> >> #include <asm/isc.h> >> #include <asm/crw.h> >> >> @@ -224,6 +226,8 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid, >> INIT_WORK(&sch->todo_work, css_sch_todo); >> sch->dev.release = &css_subchannel_release; >> device_initialize(&sch->dev); > > It might be helpful to add a comment why you use 31 bit here... @Halil, please let me know what comment you prefere here... > >> + sch->dev.coherent_dma_mask = DMA_BIT_MASK(31); >> + sch->dev.dma_mask = &sch->dev.coherent_dma_mask; >> return sch; >> >> err: >> @@ -899,6 +903,8 @@ static int __init setup_css(int nr) >> dev_set_name(&css->device, "css%x", nr); >> css->device.groups = cssdev_attr_groups; >> css->device.release = channel_subsystem_release; > > ...and 64 bit here. and here. > >> + css->device.coherent_dma_mask = DMA_BIT_MASK(64); >> + css->device.dma_mask = &css->device.coherent_dma_mask; >> >> mutex_init(&css->mutex); >> css->cssid = chsc_get_cssid(nr); > > (...) > >> @@ -1059,16 +1168,19 @@ static int __init css_bus_init(void) >> if (ret) >> goto out_unregister; >> ret = register_pm_notifier(&css_power_notifier); >> - if (ret) { >> - unregister_reboot_notifier(&css_reboot_notifier); >> - goto out_unregister; >> - } >> + if (ret) >> + goto out_unregister_rn; >> + ret = cio_dma_pool_init(); >> + if (ret) >> + goto out_unregister_rn; > > Don't you also need to unregister the pm notifier on failure here? Mmh, that was the original intention. Thanks! > > Other than that, I noticed only cosmetic issues; seems reasonable to me. > >> css_init_done = 1; >> >> /* Enable default isc for I/O subchannels. */ >> isc_register(IO_SCH_ISC); >> >> return 0; >> +out_unregister_rn: >> + unregister_reboot_notifier(&css_reboot_notifier); >> out_unregister: >> while (i-- > 0) { >> struct channel_subsystem *css = channel_subsystems[i]; > Thanks, Michael
On Mon, 3 Jun 2019 13:37:45 +0200 Cornelia Huck <cohuck@redhat.com> wrote: > On Wed, 29 May 2019 14:26:51 +0200 > Michael Mueller <mimu@linux.ibm.com> wrote: > > > From: Halil Pasic <pasic@linux.ibm.com> > > > > To support protected virtualization cio will need to make sure the > > memory used for communication with the hypervisor is DMA memory. > > > > Let us introduce one global pool for cio. > > > > Our DMA pools are implemented as a gen_pool backed with DMA pages. The > > idea is to avoid each allocation effectively wasting a page, as we > > typically allocate much less than PAGE_SIZE. > > > > Signed-off-by: Halil Pasic <pasic@linux.ibm.com> > > Reviewed-by: Sebastian Ott <sebott@linux.ibm.com> > > Signed-off-by: Michael Mueller <mimu@linux.ibm.com> > > --- > > arch/s390/Kconfig | 1 + > > arch/s390/include/asm/cio.h | 11 ++++ > > drivers/s390/cio/css.c | 120 ++++++++++++++++++++++++++++++++++++++++++-- > > 3 files changed, 128 insertions(+), 4 deletions(-) > > (...) > > > diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h > > index 1727180e8ca1..43c007d2775a 100644 > > --- a/arch/s390/include/asm/cio.h > > +++ b/arch/s390/include/asm/cio.h > > @@ -328,6 +328,17 @@ static inline u8 pathmask_to_pos(u8 mask) > > void channel_subsystem_reinit(void); > > extern void css_schedule_reprobe(void); > > > > +extern void *cio_dma_zalloc(size_t size); > > +extern void cio_dma_free(void *cpu_addr, size_t size); > > +extern struct device *cio_get_dma_css_dev(void); > > + > > +struct gen_pool; > > That forward declaration is a bit ugly... Can you explain to me what is ugly about it so I can avoid similar mistakes in the future? >I guess the alternative was > include hell? > What do you mean by include hell? I decided to use a forward declaration because the guys that include "cio.h" are not expected to require the interfaces defined in linux/genalloc.h. My motivation to do it like this was the principle of encapsulation. Regards, Halil
On Mon, 3 Jun 2019 14:09:02 +0200 Michael Mueller <mimu@linux.ibm.com> wrote: > >> @@ -224,6 +226,8 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid, > >> INIT_WORK(&sch->todo_work, css_sch_todo); > >> sch->dev.release = &css_subchannel_release; > >> device_initialize(&sch->dev); > > > > It might be helpful to add a comment why you use 31 bit here... > > @Halil, please let me know what comment you prefere here... > How about? /* * The physical addresses of some the dma structures that * can belong to a subchannel need to fit 31 bit width (examples ccw,). */ > > > >> + sch->dev.coherent_dma_mask = DMA_BIT_MASK(31); > >> + sch->dev.dma_mask = &sch->dev.coherent_dma_mask; > >> return sch; > >> > >> err: > >> @@ -899,6 +903,8 @@ static int __init setup_css(int nr) > >> dev_set_name(&css->device, "css%x", nr); > >> css->device.groups = cssdev_attr_groups; > >> css->device.release = channel_subsystem_release; > > > > ...and 64 bit here. > > and here. /* * We currently allocate notifier bits with this (using css->device * as the device argument with the DMA API), and are fine with 64 bit * addresses. */ Regards, Halil
On Mon, 3 Jun 2019 14:57:30 +0200 Halil Pasic <pasic@linux.ibm.com> wrote: > On Mon, 3 Jun 2019 14:09:02 +0200 > Michael Mueller <mimu@linux.ibm.com> wrote: > > > >> @@ -224,6 +226,8 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid, > > >> INIT_WORK(&sch->todo_work, css_sch_todo); > > >> sch->dev.release = &css_subchannel_release; > > >> device_initialize(&sch->dev); > > > > > > It might be helpful to add a comment why you use 31 bit here... > > > > @Halil, please let me know what comment you prefere here... > > > > How about? > > /* > * The physical addresses of some the dma structures that > * can belong to a subchannel need to fit 31 bit width (examples ccw,). > */ "e.g. ccw"? > > > > > > > >> + sch->dev.coherent_dma_mask = DMA_BIT_MASK(31); > > >> + sch->dev.dma_mask = &sch->dev.coherent_dma_mask; > > >> return sch; > > >> > > >> err: > > >> @@ -899,6 +903,8 @@ static int __init setup_css(int nr) > > >> dev_set_name(&css->device, "css%x", nr); > > >> css->device.groups = cssdev_attr_groups; > > >> css->device.release = channel_subsystem_release; > > > > > > ...and 64 bit here. > > > > and here. > > /* > * We currently allocate notifier bits with this (using css->device > * as the device argument with the DMA API), and are fine with 64 bit > * addresses. > */ Thanks, that makes things hopefully clearer if we look at it some time in the future ;)
On Mon, 3 Jun 2019 14:47:06 +0200 Halil Pasic <pasic@linux.ibm.com> wrote: > On Mon, 3 Jun 2019 13:37:45 +0200 > Cornelia Huck <cohuck@redhat.com> wrote: > > > On Wed, 29 May 2019 14:26:51 +0200 > > Michael Mueller <mimu@linux.ibm.com> wrote: > > > diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h > > > index 1727180e8ca1..43c007d2775a 100644 > > > --- a/arch/s390/include/asm/cio.h > > > +++ b/arch/s390/include/asm/cio.h > > > @@ -328,6 +328,17 @@ static inline u8 pathmask_to_pos(u8 mask) > > > void channel_subsystem_reinit(void); > > > extern void css_schedule_reprobe(void); > > > > > > +extern void *cio_dma_zalloc(size_t size); > > > +extern void cio_dma_free(void *cpu_addr, size_t size); > > > +extern struct device *cio_get_dma_css_dev(void); > > > + > > > +struct gen_pool; > > > > That forward declaration is a bit ugly... > > Can you explain to me what is ugly about it so I can avoid similar > mistakes in the future? > > >I guess the alternative was > > include hell? > > > > What do you mean by include hell? > > I decided to use a forward declaration because the guys that include > "cio.h" are not expected to require the interfaces defined in > linux/genalloc.h. My motivation to do it like this was the principle of > encapsulation. My personal rule-of-thumb is to include the header if it is straightforward enough (e.g. if adding a basic header is enough). If you need to include a header together with all of its friends and family, a forward declaration is probably nicer. And of course, sometimes it is simply needed.
On 03.06.19 15:34, Cornelia Huck wrote: > On Mon, 3 Jun 2019 14:57:30 +0200 > Halil Pasic <pasic@linux.ibm.com> wrote: > >> On Mon, 3 Jun 2019 14:09:02 +0200 >> Michael Mueller <mimu@linux.ibm.com> wrote: >> >>>>> @@ -224,6 +226,8 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid, >>>>> INIT_WORK(&sch->todo_work, css_sch_todo); >>>>> sch->dev.release = &css_subchannel_release; >>>>> device_initialize(&sch->dev); >>>> >>>> It might be helpful to add a comment why you use 31 bit here... >>> >>> @Halil, please let me know what comment you prefere here... >>> >> >> How about? >> >> /* >> * The physical addresses of some the dma structures that >> * can belong to a subchannel need to fit 31 bit width (examples ccw,). >> */ > > "e.g. ccw"? > >> >> >>>> >>>>> + sch->dev.coherent_dma_mask = DMA_BIT_MASK(31); >>>>> + sch->dev.dma_mask = &sch->dev.coherent_dma_mask; >>>>> return sch; >>>>> >>>>> err: >>>>> @@ -899,6 +903,8 @@ static int __init setup_css(int nr) >>>>> dev_set_name(&css->device, "css%x", nr); >>>>> css->device.groups = cssdev_attr_groups; >>>>> css->device.release = channel_subsystem_release; >>>> >>>> ...and 64 bit here. >>> >>> and here. >> >> /* >> * We currently allocate notifier bits with this (using css->device >> * as the device argument with the DMA API), and are fine with 64 bit >> * addresses. >> */ > > Thanks, that makes things hopefully clearer if we look at it some time > in the future ;) > Applied both with with requested change. Michael
On Mon, 3 Jun 2019 14:09:02 +0200 Michael Mueller <mimu@linux.ibm.com> wrote: > >> @@ -1059,16 +1168,19 @@ static int __init css_bus_init(void) > >> if (ret) > >> goto out_unregister; > >> ret = register_pm_notifier(&css_power_notifier); > >> - if (ret) { > >> - unregister_reboot_notifier(&css_reboot_notifier); > >> - goto out_unregister; > >> - } > >> + if (ret) > >> + goto out_unregister_rn; > >> + ret = cio_dma_pool_init(); > >> + if (ret) > >> + goto out_unregister_rn; > > > > Don't you also need to unregister the pm notifier on failure here? > > Mmh, that was the original intention. Thanks! I suppose we could also move cio_dma_pool_init() right before the register_reboot_notifier() call and goto out_unregister on error. Regards, Halil
On Mon, 3 Jun 2019 16:04:28 +0200 Halil Pasic <pasic@linux.ibm.com> wrote: > On Mon, 3 Jun 2019 14:09:02 +0200 > Michael Mueller <mimu@linux.ibm.com> wrote: > > > >> @@ -1059,16 +1168,19 @@ static int __init css_bus_init(void) > > >> if (ret) > > >> goto out_unregister; > > >> ret = register_pm_notifier(&css_power_notifier); > > >> - if (ret) { > > >> - unregister_reboot_notifier(&css_reboot_notifier); > > >> - goto out_unregister; > > >> - } > > >> + if (ret) > > >> + goto out_unregister_rn; > > >> + ret = cio_dma_pool_init(); > > >> + if (ret) > > >> + goto out_unregister_rn; > > > > > > Don't you also need to unregister the pm notifier on failure here? > > > > Mmh, that was the original intention. Thanks! > > I suppose we could also move cio_dma_pool_init() right before the > register_reboot_notifier() call and goto out_unregister on error. > Forget it, then we have to rollback the pool creation if the register stuff fails... Sorry for the noise. Regards, Halil
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 88d8355b7bf7..2a245b56db8b 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -191,6 +191,7 @@ config S390 select ARCH_HAS_SCALED_CPUTIME select HAVE_NMI select SWIOTLB + select GENERIC_ALLOCATOR config SCHED_OMIT_FRAME_POINTER diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h index 1727180e8ca1..43c007d2775a 100644 --- a/arch/s390/include/asm/cio.h +++ b/arch/s390/include/asm/cio.h @@ -328,6 +328,17 @@ static inline u8 pathmask_to_pos(u8 mask) void channel_subsystem_reinit(void); extern void css_schedule_reprobe(void); +extern void *cio_dma_zalloc(size_t size); +extern void cio_dma_free(void *cpu_addr, size_t size); +extern struct device *cio_get_dma_css_dev(void); + +struct gen_pool; +void *cio_gp_dma_zalloc(struct gen_pool *gp_dma, struct device *dma_dev, + size_t size); +void cio_gp_dma_free(struct gen_pool *gp_dma, void *cpu_addr, size_t size); +void cio_gp_dma_destroy(struct gen_pool *gp_dma, struct device *dma_dev); +struct gen_pool *cio_gp_dma_create(struct device *dma_dev, int nr_pages); + /* Function from drivers/s390/cio/chsc.c */ int chsc_sstpc(void *page, unsigned int op, u16 ctrl, u64 *clock_delta); int chsc_sstpi(void *page, void *result, size_t size); diff --git a/drivers/s390/cio/css.c b/drivers/s390/cio/css.c index aea502922646..b97618497848 100644 --- a/drivers/s390/cio/css.c +++ b/drivers/s390/cio/css.c @@ -20,6 +20,8 @@ #include <linux/reboot.h> #include <linux/suspend.h> #include <linux/proc_fs.h> +#include <linux/genalloc.h> +#include <linux/dma-mapping.h> #include <asm/isc.h> #include <asm/crw.h> @@ -224,6 +226,8 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid, INIT_WORK(&sch->todo_work, css_sch_todo); sch->dev.release = &css_subchannel_release; device_initialize(&sch->dev); + sch->dev.coherent_dma_mask = DMA_BIT_MASK(31); + sch->dev.dma_mask = &sch->dev.coherent_dma_mask; return sch; err: @@ -899,6 +903,8 @@ static int __init setup_css(int nr) dev_set_name(&css->device, "css%x", nr); css->device.groups = cssdev_attr_groups; css->device.release = channel_subsystem_release; + css->device.coherent_dma_mask = DMA_BIT_MASK(64); + css->device.dma_mask = &css->device.coherent_dma_mask; mutex_init(&css->mutex); css->cssid = chsc_get_cssid(nr); @@ -1018,6 +1024,109 @@ static struct notifier_block css_power_notifier = { .notifier_call = css_power_event, }; +#define CIO_DMA_GFP (GFP_KERNEL | __GFP_ZERO) +static struct gen_pool *cio_dma_pool; + +/* Currently cio supports only a single css */ +struct device *cio_get_dma_css_dev(void) +{ + return &channel_subsystems[0]->device; +} + +struct gen_pool *cio_gp_dma_create(struct device *dma_dev, int nr_pages) +{ + struct gen_pool *gp_dma; + void *cpu_addr; + dma_addr_t dma_addr; + int i; + + gp_dma = gen_pool_create(3, -1); + if (!gp_dma) + return NULL; + for (i = 0; i < nr_pages; ++i) { + cpu_addr = dma_alloc_coherent(dma_dev, PAGE_SIZE, &dma_addr, + CIO_DMA_GFP); + if (!cpu_addr) + return gp_dma; + gen_pool_add_virt(gp_dma, (unsigned long) cpu_addr, + dma_addr, PAGE_SIZE, -1); + } + return gp_dma; +} + +static void __gp_dma_free_dma(struct gen_pool *pool, + struct gen_pool_chunk *chunk, void *data) +{ + size_t chunk_size = chunk->end_addr - chunk->start_addr + 1; + + dma_free_coherent((struct device *) data, chunk_size, + (void *) chunk->start_addr, + (dma_addr_t) chunk->phys_addr); +} + +void cio_gp_dma_destroy(struct gen_pool *gp_dma, struct device *dma_dev) +{ + if (!gp_dma) + return; + /* this is qite ugly but no better idea */ + gen_pool_for_each_chunk(gp_dma, __gp_dma_free_dma, dma_dev); + gen_pool_destroy(gp_dma); +} + +static int cio_dma_pool_init(void) +{ + /* No need to free up the resources: compiled in */ + cio_dma_pool = cio_gp_dma_create(cio_get_dma_css_dev(), 1); + if (!cio_dma_pool) + return -ENOMEM; + return 0; +} + +void *cio_gp_dma_zalloc(struct gen_pool *gp_dma, struct device *dma_dev, + size_t size) +{ + dma_addr_t dma_addr; + unsigned long addr; + size_t chunk_size; + + addr = gen_pool_alloc(gp_dma, size); + while (!addr) { + chunk_size = round_up(size, PAGE_SIZE); + addr = (unsigned long) dma_alloc_coherent(dma_dev, + chunk_size, &dma_addr, CIO_DMA_GFP); + if (!addr) + return NULL; + gen_pool_add_virt(gp_dma, addr, dma_addr, chunk_size, -1); + addr = gen_pool_alloc(gp_dma, size); + } + return (void *) addr; +} + +void cio_gp_dma_free(struct gen_pool *gp_dma, void *cpu_addr, size_t size) +{ + if (!cpu_addr) + return; + memset(cpu_addr, 0, size); + gen_pool_free(gp_dma, (unsigned long) cpu_addr, size); +} + +/* + * Allocate dma memory from the css global pool. Intended for memory not + * specific to any single device within the css. The allocated memory + * is not guaranteed to be 31-bit addressable. + * + * Caution: Not suitable for early stuff like console. + */ +void *cio_dma_zalloc(size_t size) +{ + return cio_gp_dma_zalloc(cio_dma_pool, cio_get_dma_css_dev(), size); +} + +void cio_dma_free(void *cpu_addr, size_t size) +{ + cio_gp_dma_free(cio_dma_pool, cpu_addr, size); +} + /* * Now that the driver core is running, we can setup our channel subsystem. * The struct subchannel's are created during probing. @@ -1059,16 +1168,19 @@ static int __init css_bus_init(void) if (ret) goto out_unregister; ret = register_pm_notifier(&css_power_notifier); - if (ret) { - unregister_reboot_notifier(&css_reboot_notifier); - goto out_unregister; - } + if (ret) + goto out_unregister_rn; + ret = cio_dma_pool_init(); + if (ret) + goto out_unregister_rn; css_init_done = 1; /* Enable default isc for I/O subchannels. */ isc_register(IO_SCH_ISC); return 0; +out_unregister_rn: + unregister_reboot_notifier(&css_reboot_notifier); out_unregister: while (i-- > 0) { struct channel_subsystem *css = channel_subsystems[i];