Re: [PATCH v4 13/21] remoterproc: qcom: refactor to leverage exported minidump symbol

[Pavan Kondeti]

On Thu, Jun 29, 2023 at 02:50:34PM +0530, Mukesh Ojha wrote:
> 
> 
> On 6/28/2023 9:21 PM, Pavan Kondeti wrote:
> > On Wed, Jun 28, 2023 at 06:04:40PM +0530, Mukesh Ojha wrote:
> > > -static int qcom_add_minidump_segments(struct rproc *rproc, struct minidump_subsystem *subsystem,
> > > -			void (*rproc_dumpfn_t)(struct rproc *rproc, struct rproc_dump_segment *segment,
> > > -				void *dest, size_t offset, size_t size))
> > > +void qcom_minidump(struct rproc *rproc, unsigned int minidump_id,
> > > +		void (*rproc_dumpfn_t)(struct rproc *rproc,
> > > +		struct rproc_dump_segment *segment, void *dest, size_t offset,
> > > +		size_t size))
> > >   {
> > > -	struct minidump_region __iomem *ptr;
> > > -	struct minidump_region region;
> > > -	int seg_cnt, i;
> > >   	dma_addr_t da;
> > >   	size_t size;
> > > +	int seg_cnt;
> > >   	char *name;
> > > +	void *ptr;
> > > +	int ret;
> > > +	int i;
> > >   	if (WARN_ON(!list_empty(&rproc->dump_segments))) {
> > >   		dev_err(&rproc->dev, "dump segment list already populated\n");
> > > -		return -EUCLEAN;
> > > +		return;
> > >   	}
> > > -	seg_cnt = le32_to_cpu(subsystem->region_count);
> > > -	ptr = ioremap((unsigned long)le64_to_cpu(subsystem->regions_baseptr),
> > > -		      seg_cnt * sizeof(struct minidump_region));
> > > +	ptr = qcom_ss_md_mapped_base(minidump_id, &seg_cnt);
> > >   	if (!ptr)
> > > -		return -EFAULT;
> > > +		return;
> > >   	for (i = 0; i < seg_cnt; i++) {
> > > -		memcpy_fromio(&region, ptr + i, sizeof(region));
> > > -		if (le32_to_cpu(region.valid) == MINIDUMP_REGION_VALID) {
> > > -			name = kstrndup(region.name, MAX_REGION_NAME_LENGTH - 1, GFP_KERNEL);
> > > -			if (!name) {
> > > -				iounmap(ptr);
> > > -				return -ENOMEM;
> > > -			}
> > > -			da = le64_to_cpu(region.address);
> > > -			size = le64_to_cpu(region.size);
> > > -			rproc_coredump_add_custom_segment(rproc, da, size, rproc_dumpfn_t, name);
> > > +		ret = qcom_ss_valid_segment_info(ptr, i, &name, &da, &size);
> > > +		if (ret < 0) {
> > > +			iounmap(ptr);
> > > +			dev_err(&rproc->dev,
> > > +				"Failed with error: %d while adding minidump entries\n",
> > > +				ret);
> > > +			goto clean_minidump;
> > >   		}
> > > -	}
> > > -
> > > -	iounmap(ptr);
> > > -	return 0;
> > > -}
> > > -
> > > -void qcom_minidump(struct rproc *rproc, unsigned int minidump_id,
> > > -		void (*rproc_dumpfn_t)(struct rproc *rproc,
> > > -		struct rproc_dump_segment *segment, void *dest, size_t offset,
> > > -		size_t size))
> > > -{
> > > -	int ret;
> > > -	struct minidump_subsystem *subsystem;
> > > -	struct minidump_global_toc *toc;
> > > -
> > > -	/* Get Global minidump ToC*/
> > > -	toc = qcom_smem_get(QCOM_SMEM_HOST_ANY, SBL_MINIDUMP_SMEM_ID, NULL);
> > > -
> > > -	/* check if global table pointer exists and init is set */
> > > -	if (IS_ERR(toc) || !toc->status) {
> > > -		dev_err(&rproc->dev, "Minidump TOC not found in SMEM\n");
> > > -		return;
> > > -	}
> > > -	/* Get subsystem table of contents using the minidump id */
> > > -	subsystem = &toc->subsystems[minidump_id];
> > > -
> > > -	/**
> > > -	 * Collect minidump if SS ToC is valid and segment table
> > > -	 * is initialized in memory and encryption status is set.
> > > -	 */
> > > -	if (subsystem->regions_baseptr == 0 ||
> > > -	    le32_to_cpu(subsystem->status) != 1 ||
> > > -	    le32_to_cpu(subsystem->enabled) != MINIDUMP_SS_ENABLED ||
> > > -	    le32_to_cpu(subsystem->encryption_status) != MINIDUMP_SS_ENCR_DONE) {
> > > -		dev_err(&rproc->dev, "Minidump not ready, skipping\n");
> > > -		return;
> > > +		/* if it is a valid segment */
> > > +		if (!ret)
> > > +			rproc_coredump_add_custom_segment(rproc, da, size,
> > > +							  rproc_dumpfn_t, name);
> > >   	}
> > > -	ret = qcom_add_minidump_segments(rproc, subsystem, rproc_dumpfn_t);
> > > -	if (ret) {
> > > -		dev_err(&rproc->dev, "Failed with error: %d while adding minidump entries\n", ret);
> > > -		goto clean_minidump;
> > > -	}
> > > +	iounmap(ptr);
> > >   	rproc_coredump_using_sections(rproc);
> > > +
> > >   clean_minidump:
> > >   	qcom_minidump_cleanup(rproc);
> > >   }
> > 
> > I like the idea of moving minidump pieces to drivers/soc/qcom/*minidump*.
> > 
> > Is it possible to accept one function callback from remoteproc and do
> > all of this in one function exported by minidump?
> > 
> > qcom_ss_valid_segment_info() seems a low level function to be exported..
> 
> It was ending up with circular dependency due to
> rproc_coredump_add_custom_segment()
> 
> 
> rproc_coredump => qcom_common => qcom_minidump_smem => rproc_coredump
> 

Where is the circular dependency here? Any API accepting callback would
end up like below

caller -> API(callback) -> (*callback) -> [in caller code ...]

May be I am missing something here.

AFAICS, the minidump could do everything except it does not know how to
create a segment and add it to rproc::dump_segments list. Is it not
possible to write a minidump API that accepts a callback and a list
head? we may not probably re-use rproc_coredump_add_custom_segment() as
is, but some refactoring allows/covers both cases..Pls give it a
thought.

This way we don't need to create an internal API like below

int qcom_ss_valid_segment_info(void *ptr, int i, char **name, dma_addr_t
*da, size_t *size)

Thanks,
Pavan