Re: [PATCH v3 10/12] ppc64/kexec_file: prepare elfcore header for crashing kernel

[Hari Bathini <hbathini@linux.ibm.com>]

On 16/07/20 7:52 am, Thiago Jung Bauermann wrote:
> 
> Hari Bathini <hbathini@linux.ibm.com> writes:
> 
>>  /**
>> + * get_crash_memory_ranges - Get crash memory ranges. This list includes
>> + *                           first/crashing kernel's memory regions that
>> + *                           would be exported via an elfcore.
>> + * @mem_ranges:              Range list to add the memory ranges to.
>> + *
>> + * Returns 0 on success, negative errno on error.
>> + */
>> +static int get_crash_memory_ranges(struct crash_mem **mem_ranges)
>> +{
>> +	struct memblock_region *reg;
>> +	struct crash_mem *tmem;
>> +	int ret;
>> +
>> +	for_each_memblock(memory, reg) {
>> +		u64 base, size;
>> +
>> +		base = (u64)reg->base;
>> +		size = (u64)reg->size;
>> +
>> +		/* Skip backup memory region, which needs a separate entry */
>> +		if (base == BACKUP_SRC_START) {
>> +			if (size > BACKUP_SRC_SIZE) {
>> +				base = BACKUP_SRC_END + 1;
>> +				size -= BACKUP_SRC_SIZE;
>> +			} else
>> +				continue;
>> +		}
>> +
>> +		ret = add_mem_range(mem_ranges, base, size);
>> +		if (ret)
>> +			goto out;
>> +
>> +		/* Try merging adjacent ranges before reallocation attempt */
>> +		if ((*mem_ranges)->nr_ranges == (*mem_ranges)->max_nr_ranges)
>> +			sort_memory_ranges(*mem_ranges, true);
>> +	}
>> +
>> +	/* Reallocate memory ranges if there is no space to split ranges */
>> +	tmem = *mem_ranges;
>> +	if (tmem && (tmem->nr_ranges == tmem->max_nr_ranges)) {
>> +		tmem = realloc_mem_ranges(mem_ranges);
>> +		if (!tmem)
>> +			goto out;
>> +	}
>> +
>> +	/* Exclude crashkernel region */
>> +	ret = crash_exclude_mem_range(tmem, crashk_res.start, crashk_res.end);
>> +	if (ret)
>> +		goto out;
>> +
>> +	ret = add_rtas_mem_range(mem_ranges);
>> +	if (ret)
>> +		goto out;
>> +
>> +	ret = add_opal_mem_range(mem_ranges);
>> +	if (ret)
>> +		goto out;
> 
> Maybe I'm confused, but don't you add the RTAS and OPAL regions as
> usable memory for the crashkernel? In that case they shouldn't show up
> in the core file.

kexec-tools does the same thing. I am not endorsing it but I was trying to stay
in parity to avoid breaking any userspace tools/commands. But as you rightly
pointed, this is NOT right. The right thing to do, to get the rtas/opal data at
the time of crash, is to have a backup region for them just like we have for
the first 64K memory. I was hoping to do that later.

Will check how userspace tools respond to dropping these regions. If that makes
the tools unhappy, will retain the regions with a FIXME. Sorry about the confusion.

Thanks
Hari