From mboxrd@z Thu Jan  1 00:00:00 1970
From: hch@infradead.org (Christoph Hellwig)
Date: Thu, 21 Jun 2018 08:24:07 -0700
Subject: [PATCH] nvme: use __GFP_NOWARN for iod allocation
In-Reply-To: <d9fcda4e-9ed2-5001-ef38-61f7c0247cfa@kernel.dk>
References: <b51cce4d-d0c9-fb4e-d5ac-2e48e23e18f7@kernel.dk>
 <20180620204310.GA24840@localhost.localdomain>
 <34e64832-c345-3a70-401f-1915b5bb4af1@kernel.dk>
 <20180620223035.GA25196@localhost.localdomain>
 <7b6f885f-f605-1bfd-7a7e-2fce8cbde6c2@kernel.dk>
 <20180621071056.GB21304@infradead.org>
 <dc92becd-403d-9c15-e5bb-5e0c3b5d8a8c@kernel.dk>
 <20180621150139.GA22804@infradead.org>
 <cb33de73-6b03-d4a8-ca73-dc094f83f742@kernel.dk>
 <d9fcda4e-9ed2-5001-ef38-61f7c0247cfa@kernel.dk>
Message-ID: <20180621152407.GA29947@infradead.org>

I guess we can live with it for now with a big comment

>  	ctrl->cntlid = le16_to_cpup(&id->cntlid);
>  	if (id->mdts)
>  		max_hw_sectors = 1 << (id->mdts + page_shift - 9);
> -	else
> +	else if (ctrl->ops->flags & NVME_F_FABRICS)
>  		max_hw_sectors = UINT_MAX;
> +	else
> +		max_hw_sectors = NVME_MAX_KB_SZ << 1;

Can this I think is better done by setting ctrl->max_hw_sectors
in the PCIe driver (similar to what the other drivers already do)
and then rely on the min_not_zero below to factor in MDTS.

> +/*
> + * These can be higher, but we need to ensure that any command doesn't
> + * require an sg allocation that needs more than a page of data.
> + */
> +#define NVME_MAX_KB_SZ	4096

I think this should move the pci.c

> +#define NVME_MAX_SEGS	127

And this could use a good explanation.  Especially as I think
that we shouldn't apply the limit for fabrics.  So instead we
might just add a new ctrl->max_segments field, which PCI sets,
and which then is min() with the otherwise calculated value.