On 10/14/24 4:54 PM, Ritesh Harjani (IBM) wrote:

Madhavan Srinivasan [off-list ref] writes:

quoted

On 10/11/24 8:30 PM, Ritesh Harjani (IBM) wrote:

quoted

We anyway don't use any return values from fadump_cma_init(). Since
fadump_reserve_mem() from where fadump_cma_init() gets called today,
already has the required checks.
This patch makes this function return type as void. Let's also handle
extra cases like return if fadump_supported is false or dump_active, so
that in later patches we can call fadump_cma_init() separately from
setup_arch().

Usually patches to this file are posted with title format of

powerpc/fadump:<>

yes. I guess it is good to do it that way (I might have missed it)
Although commit history of oldest few patches to fadump shows..

ebaeb5ae2437 fadump: Convert firmware-assisted cpu state dump data into elf notes.
2df173d9e85d fadump: Initialize elfcore header and add PT_LOAD program headers.
3ccc00a7e04f fadump: Register for firmware assisted dump.
eb39c8803d0e fadump: Reserve the memory for firmware assisted dump.

quoted

Patchset looks fine to me. 

Reviewed-by: Madhavan Srinivasan <maddy@linux.ibm.com> for the series.

quoted

quoted

Acked-by: Hari Bathini <hbathini@linux.ibm.com>
Signed-off-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
---
v2 -> v3: Separated the series into 2 as discussed in v2.
[v2]: https://lore.kernel.org/linuxppc-dev/cover.1728585512.git.ritesh.list@gmail.com/ (local)

 arch/powerpc/kernel/fadump.c | 23 +++++++++--------------
 1 file changed, 9 insertions(+), 14 deletions(-)

diff --git a/arch/powerpc/kernel/fadump.c b/arch/powerpc/kernel/fadump.c
index a612e7513a4f..162327d66982 100644
--- a/arch/powerpc/kernel/fadump.c
+++ b/arch/powerpc/kernel/fadump.c

@@ -78,27 +78,23 @@ static struct cma *fadump_cma;
  * But for some reason even if it fails we still have the memory reservation
  * with us and we can still continue doing fadump.
  */
-static int __init fadump_cma_init(void)
+static void __init fadump_cma_init(void)
 {
 	unsigned long long base, size;
 	int rc;

-	if (!fw_dump.fadump_enabled)
-		return 0;
-
+	if (!fw_dump.fadump_supported || !fw_dump.fadump_enabled ||
+			fw_dump.dump_active)
+		return;

Is these checks even needed here? fadump_reserve_mem() checked for all
these already, also dont see any other caller for fadump_cma_init(). 

In the next patch we will move fadump_cma_init() call from within
fadump_reserve_mem() to setup_arch(). Hence we need these extra checks
in fadump_cma_init() as well. I mentioned the same in the commit msg of
this patch too.

quoted

quoted

 	/*
 	 * Do not use CMA if user has provided fadump=nocma kernel parameter.
-	 * Return 1 to continue with fadump old behaviour.
 	 */
-	if (fw_dump.nocma)
-		return 1;
+	if (fw_dump.nocma || !fw_dump.boot_memory_size)
+		return;

 	base = fw_dump.reserve_dump_area_start;
 	size = fw_dump.boot_memory_size;

-	if (!size)
-		return 0;

So this is the only place where we return 0, which in turn will make the
"ret" in fadump_reserve_mem() as zero forcing to call reserve_crashkernel()
in early_init_devtree().

we are removing it, becos we know "size" here will never be zero?

yes. Because we already check if boot_memory_size is less than
bootmem_min in fadump_reserve_mem(). If it is less, then we fail and
disable fadump (fadump_enabled = 0).

So then there is no need to check for !boot_memory_size in here.

fadump_reseve_mem( ) {
<...>
	if (!fw_dump.dump_active) {
		fw_dump.boot_memory_size =
			PAGE_ALIGN(fadump_calculate_reserve_size());

		bootmem_min = fw_dump.ops->fadump_get_bootmem_min();
		if (fw_dump.boot_memory_size < bootmem_min) {
			pr_err("Can't enable fadump with boot memory size (0x%lx) less than 0x%llx\n",
			       fw_dump.boot_memory_size, bootmem_min);
			goto error_out;
		}
    <...>    
    }
<...>
error_out:
	fw_dump.fadump_enabled = 0;
	fw_dump.reserve_dump_area_size = 0;
	return 0;
}


Thanks for the review!
-ritesh