Re: [tarantool-patches] [PATCH v3 06/14] txn: get rid of fiber_gc from txn_rollback

[Vladimir Davydov <vdavydov.dev@gmail.com>]

On Sun, Jun 09, 2019 at 11:44:35PM +0300, Georgy Kirichenko wrote:
> Don't touch a fiber gc storage on a transaction rollback explicitly.
> This relaxes dependencies between fiber and transaction life cycles.
> 
> Prerequisites: #1254
> ---
>  src/box/applier.cc     |  8 +++++---
>  src/box/box.cc         | 14 +++++++++-----
>  src/box/call.c         | 22 ++++++++++++++++------
>  src/box/memtx_engine.c |  3 ++-
>  src/box/txn.c          | 35 +++++++++++++++++------------------
>  src/box/txn.h          |  8 ++++++--
>  src/box/vy_scheduler.c | 10 +++++++---
>  7 files changed, 62 insertions(+), 38 deletions(-)
> 
> diff --git a/src/box/applier.cc b/src/box/applier.cc
> index e3203a4c8..5a92f6109 100644
> --- a/src/box/applier.cc
> +++ b/src/box/applier.cc
> @@ -190,7 +190,7 @@ apply_initial_join_row(struct xrow_header *row)
>  	fiber_gc();
>  	return rc;
>  rollback:
> -	txn_rollback();
> +	txn_rollback(txn);
>  	return -1;
>  }
>  
> @@ -422,7 +422,8 @@ applier_join(struct applier *applier)
>  			if (txn == NULL)
>  				diag_raise();
>  			if (apply_row(&row) != 0) {
> -				txn_rollback();
> +				txn_rollback(txn);
> +				fiber_gc();
>  				diag_raise();
>  			}
>  			if (txn_commit(txn) != 0)
> @@ -625,7 +626,8 @@ applier_apply_tx(struct stailq *rows)
>  	return txn_commit(txn);
>  
>  rollback:
> -	txn_rollback();
> +	txn_rollback(txn);
> +	fiber_gc();
>  	return -1;
>  }

As I wrote earlier, I'm pretty convinced we don't need to call
fiber_gc() after txn_commit/rollback at all. Please consider
removing it in a separate patch.