Re: [PATCH REBASED v5 1/2] spapr: nvdimm: Implement H_SCM_FLUSH hcall

[Shivaprasad G Bhat <sbhat@linux.ibm.com>]

Hi David,

Thanks for comments. Sorry about the delay. Replies inline.

On 9/21/21 11:53, David Gibson wrote:
> On Wed, Jul 07, 2021 at 09:57:21PM -0500, Shivaprasad G Bhat wrote:
>> The patch adds support for the SCM flush hcall for the nvdimm devices.
>> To be available for exploitation by guest through the next patch.
>>
>> The hcall expects the semantics such that the flush to return
>> with one of H_LONG_BUSY when the operation is expected to take longer
>> time along with a continue_token. The hcall to be called again providing
>> the continue_token to get the status. So, all fresh requests are put into
>> a 'pending' list and flush worker is submitted to the thread pool. The
>> thread pool completion callbacks move the requests to 'completed' list,
>> which are cleaned up after reporting to guest in subsequent hcalls t

<snip>

>> @@ -30,6 +31,7 @@
>>   #include "hw/ppc/fdt.h"
>>   #include "qemu/range.h"
>>   #include "hw/ppc/spapr_numa.h"
>> +#include "block/thread-pool.h"
>>   
>>   /* DIMM health bitmap bitmap indicators. Taken from kernel's papr_scm.c */
>>   /* SCM device is unable to persist memory contents */
>> @@ -375,6 +377,243 @@ static target_ulong h_scm_bind_mem(PowerPCCPU *cpu, SpaprMachineState *spapr,
>>       return H_SUCCESS;
>>   }
>>   
>> +static uint64_t flush_token;
> 
> Better to put this in the machine state structure than a global.

Moved it to device state itself as suggested, the states list is per 
device now.

> 
>> +static int flush_worker_cb(void *opaque)
>> +{
>> +    int ret = H_SUCCESS;
>> +    SpaprNVDIMMDeviceFlushState *state = opaque;
>> +
>> +    /* flush raw backing image */
>> +  

<snip>

>> +             !QLIST_EMPTY(&spapr->completed_flush_states));
>> +}
>> +
>> +static int spapr_nvdimm_post_load(void *opaque, int version_id)
>> +{
>> +    SpaprMachineState *spapr = (SpaprMachineState *)opaque;
>> +    SpaprNVDIMMDeviceFlushState *state, *next;
>> +    PCDIMMDevice *dimm;
>> +    HostMemoryBackend *backend = NULL;
>> +    ThreadPool *pool = aio_get_thread_pool(qemu_get_aio_context());
>> +    SpaprDrc *drc;
>> +
>> +    QLIST_FOREACH_SAFE(state, &spapr->completed_flush_states, node, next) {
> 
> I don't think you need FOREACH_SAFE here.  You're not removing entries
> from the loop body.  If you're trying to protect against concurrent
> removals, I don't think FOREACH_SAFE is sufficient, you'll need an
> actual lock (but I think it's already protected by the BQL).

Changing here, below and also at spapr_nvdimm_get_flush_status() while 
traversing the pending list. Verified all these invocations are called 
with BQL.

> 
>> +        if (flush_token < state->continue_token) {
>> +            flush_token = state->continue_token;
>> +        }
>> +    }
>> +
>> +    QLIST_FOREACH_SAFE(state, &spapr->pending_flush_states, node, next) {
> 
> Sane comments here.
> 
>> +        if (flush_token < state->continue_token) {
>> +            flush_token = state->continue_token;
>> +        }
>> +
>> +        drc = spapr_drc_by_index(state->drcidx);
>> +        dimm = PC_DIMM(drc->dev);
>> +        backend = MEMORY_BACKEND(dimm->hostmem);
>> +        state->backend_fd = memory_region_get_fd(&backend->mr);
>> +
>> +        thread_pool_submit_aio(pool, flush_worker_cb, state,
>> +                               spapr_nvdimm_flush_completion_cb, state);
>> +    }
>> +
>> +    return 0;
>> +}
>> +
>> +const VMStateDescription vmstate_spapr_nvdimm_states = {
>> +    .name = "spapr_nvdimm_states",
>> +    .version_id = 1,
>> +    .minimum_version_id = 1,
>> +    .needed = spapr_nvdimm_states_needed,
>> +    .post_load = spapr_nvdimm_post_load,
>> +    .fields = (VMStateField[]) {
>> +        VMSTATE_QLIST_V(completed_flush_states, SpaprMachineState, 1,
>> +                        vmstate_spapr_nvdimm_flush_state,
>> +                        SpaprNVDIMMDeviceFlushState, node),
>> +        VMSTATE_QLIST_V(pending_flush_states, SpaprMachineState, 1,
>> +                        vmstate_spapr_nvdimm_flush_state,
>> +                        SpaprNVDIMMDeviceFlushState, node),
>> +        VMSTATE_END_OF_LIST()
>> +    },
>> +};
>> +
>> +/*
>> + * Assign a token and reserve it for the new flush state.
>> + */
>> +static SpaprNVDIMMDeviceFlushState *spapr_nvdimm_init_new_flush_state(
>> +                                                      SpaprMachineState *spapr)
>> +{
>> +    SpaprNVDIMMDeviceFlushState *state;
>> +
>> +    state = g_malloc0(sizeof(*state));
>> +
>> +    flush_token++;
>> +    /* Token zero is presumed as no job pending. Handle the overflow to zero */
>> +    if (flush_token == 0) {
>> +        flush_token++;
> 
> Hmm... strictly speaking, this isn't safe.  It's basically never going
> to happen in practice, but in theory there's nothing preventing
> continue_token 1 still being outstanding when the flush_token counter
> overflows.
> 
> Come to think of it, since it's a uint64_t, I think an actual overflow
> is also never going to happen in practice.  Maybe we should just
> assert() on overflow, and fix it in the unlikely event that we ever
> discover a case where it could happen.

Have added the assert on overflow.

> 
>> +    }
>> +    state->continue_token = flush_token;
>> +
>> +    QLIST_INSERT_HEAD(&spapr->pending_flush_states, state, node);
>> +
>> +    return state;
>> +}
>> +
>> +/*
>> + *

Thanks!