| Message ID | 1445400844-24370-1-git-send-email-joel@jms.id.au |
|---|---|
| State | Accepted |
| Headers | show |
Hi Joel, > On a system the PRD daemon was starting without a PNOR path, and it was > hard to diagnose what had happened. These messages should help in that > case. Acked-by: Jeremy Kerr <jk@ozlabs.org> Bonus points: have opal-prd find the PNOR device automatically. :) Cheers, Jeremy
On 10/21/2015 07:12 AM, Jeremy Kerr wrote: > Hi Joel, > >> On a system the PRD daemon was starting without a PNOR path, and it was >> hard to diagnose what had happened. These messages should help in that >> case. > > Acked-by: Jeremy Kerr <jk@ozlabs.org> > > Bonus points: have opal-prd find the PNOR device automatically. :) There should be code in gard doing it. Btw, couldn't we use the get_dev_mtd() routine from external/common/ in the opal-prd daemon and in the gard command ? This looks like common code. C.
On Thu, Oct 22, 2015 at 12:06 AM, Cédric Le Goater <clg@fr.ibm.com> wrote: >>> On a system the PRD daemon was starting without a PNOR path, and it was >>> hard to diagnose what had happened. These messages should help in that >>> case. >> >> Acked-by: Jeremy Kerr <jk@ozlabs.org> >> >> Bonus points: have opal-prd find the PNOR device automatically. :) > > There should be code in gard doing it. > > Btw, couldn't we use the get_dev_mtd() routine from external/common/ in > the opal-prd daemon and in the gard command ? This looks like common code. Good idea, thanks both of you for the suggestion. This code was written pre-gard tool so it would be good to share some code. Jeremy, we're happy to pull in the common code to opal-prd? Joel
Hi Joel, > Good idea, thanks both of you for the suggestion. This code was > written pre-gard tool so it would be good to share some code. > > Jeremy, we're happy to pull in the common code to opal-prd? Yep, definitely. As long as it builds :D Cheers, Jeremy
Joel Stanley <joel@jms.id.au> writes: > On a system the PRD daemon was starting without a PNOR path, and it was > hard to diagnose what had happened. These messages should help in that > case. > > Signed-off-by: Joel Stanley <joel@jms.id.au> Thanks, applied as of 22f0fc39a7f9
diff --git a/external/opal-prd/opal-prd.c b/external/opal-prd/opal-prd.c index 83c5cb5..0004b3d 100644 --- a/external/opal-prd/opal-prd.c +++ b/external/opal-prd/opal-prd.c @@ -1499,6 +1499,8 @@ static int run_prd_daemon(struct opal_prd_ctx *ctx) /* log to syslog */ pr_log_daemon_init(); + pr_debug("CTRL: Starting PRD daemon\n"); + ctx->fd = -1; ctx->socket = -1; @@ -1560,6 +1562,8 @@ static int run_prd_daemon(struct opal_prd_ctx *ctx) pr_log(LOG_ERR, "PNOR: Failed to open pnor: %m"); goto out_close; } + } else { + pr_log(LOG_WARNING,"PNOR: No path provided, access will fail\n"); } ipmi_init(ctx); @@ -1585,6 +1589,7 @@ static int run_prd_daemon(struct opal_prd_ctx *ctx) rc = 0; out_close: + pr_debug("CTRL: stopping PRD daemon\n"); pnor_close(&ctx->pnor); if (ctx->fd != -1) close(ctx->fd);
On a system the PRD daemon was starting without a PNOR path, and it was hard to diagnose what had happened. These messages should help in that case. Signed-off-by: Joel Stanley <joel@jms.id.au> --- external/opal-prd/opal-prd.c | 5 +++++ 1 file changed, 5 insertions(+)